CADDi
CADDi 的工业 AI 平台凭专有图纸智能形成技术差异化,并获得大型企业验证;但其高估值仍建立在不透明的私营经营指标之上。
覆盖范围与披露说明
公开披露与第三方报道口径不一;经审计的非公开经营指标仍无法获取。
封面要素
公司概况
CADDi 创立于日本,设芝加哥与东京双总部,是一家制造业 AI 数据平台,将离散制造企业的传统图纸、CAD 文件与采购记录转化为可检索的企业级工作流。
- 创始人
- Yushiro Kato, Aki Kobashi
- 创立地点
- Tokyo, Japan
- 总部
- Chicago, Illinois and Tokyo, Japan
- 产品
- 制造业 AI 数据平台,涵盖 CADDi Drawer、CADDi Explorer、CADDi Agent 以及用于图纸检索、采购情报与制造决策自动化的工作流应用。
- 客户
- 汽车、重型机械与精密制造领域的离散制造企业
- 商业模式
- 企业级 SaaS 订阅,辅以实施服务及采购业务支持
- 阶段
- Series D
- 融资情况
- 2026 年 9 月完成 $114 million Series D 轮融资,投后估值 $1.2 billion;累计股权融资 $234 million。
执行摘要
主要优势
- 围绕传统制造数据建立的专有图纸智能壁垒
- 日本制造业广泛采用,提供有力的企业级验证
- 头部资方背书,并向 Explorer、Agent 及工作流模块持续延展产品矩阵
主要风险
- 收入、利润率和客户留存数据披露仍不透明
- 重度介入的部署模式带来偏服务、人力密集的成本结构
- 传统 PLM 与 ERP 巨头可能直接捆绑同类竞品 AI 功能
未决问题
- 经审计的收入、ARR 以及软件与服务的具体拆分仍未披露
- 日本本土以外市场的毛利率与净留存率(NRR)尚未得到验证
- 北美客户拓展进展与续约稳定性仍需持续尽调
目录
01公司概况
1.1 企业定位、使命与全球布局
CADDi 于 2017 年在东京创立,核心使命旨在打破工业供应链固有的物理与数据瓶颈,释放制造业潜能。为推进跨太平洋业务扩张并深度服务西方工业制造商,公司设立伊利诺伊州芝加哥与日本东京双总部架构。历经九年稳健运营,CADDi 将全球企业级业务拓展至 22 个国家,覆盖重型工业机械、汽车及半导体设备客户。 旗下核心软件产品迅速获客;日本 Nikkei 225 成分股中的离散制造企业,已有超过 50% 部署该软件。为服务跨区域客户,CADDi 将全球全职员工扩充至 910 余人,并在北美和亚洲组建本地销售、客户成功及解决方案架构团队。 [CO001, CO002, CO011, CO012]
| 指标 | 数值 / 状态 | 截至日期 | 置信度 | 证据缺口 |
|---|---|---|---|---|
| 估值 | $1.2B 投后 | 2026-09-16 | 高 | 无;Series D 融资公告已确认 |
| 累计股权融资 | 累计 $234M | 2026-09-16 | 高 | 无;公司与投资方披露均已证实 |
| 营业收入 / ARR | 未披露 | 2026-09-26 | 低 | 公司未披露经审计的 ARR 及 GAAP 营业收入 |
| 员工规模 | 910+ 人 | 2026-09-16 | 高 | 无;官网及媒体报道均有披露 |
| 客户覆盖 | 22 个国家;50%+ 日经 225 企业 | 2026-08-01 | 高 | 2024 年转型后未披露具体企业客户数 |
关键业绩指标汇总自 CADDi 官方公告及 2026 年 9 月经核实的 Series D 融资新闻稿。
[CO005, CO006, CO011, CO012]1.2 创始人、高管团队与治理架构
CADDi 领导层兼具工业供应链管理咨询专长与硅谷软硬件工程经验。联合创始人兼 CEO Yushiro Kato 曾就职于 McKinsey & Company 主导制造业采购优化项目,为公司早期发展奠定运营基础与商业网络。联合创始人兼 CTO Aki Kobashi 拥有 Stanford University 电气工程学士及硕士学位,曾在美国航空航天领域研发大规模卫星图像处理系统,并在 Apple 负责 iPhone、AirPods 等旗舰消费硬件的产品工程研发。高管团队还包括首席财务官 Ryota Haga(统筹全球资本运作与机构投资者银团)以及工程副总裁 Chris Cope(主管北美与日本的分布式软件工程)。尽管团队对 Kato 与 Kobashi 的关键人物依赖依然显著,但芝加哥高管团队的组建正赋予公司更强的日常运营自主权。 [CO001, CO003, CO004, CO012]
| 高管姓名 | 职务 | 背景与过往履历 | 创始人与市场契合度及行业覆盖 | 关键人依赖度 |
|---|---|---|---|---|
| Yushiro Kato | 联合创始人兼 CEO | 曾任麦肯锡(McKinsey & Company)顾问,专注制造业采购优化。 | 对制造业供应链具备深厚行业理解,掌握日本及北美高管网络人脉。 | 关键;核心战略设计者与公司对外发声人。 |
| Aki Kobashi | 联合创始人兼 CTO | 斯坦福大学电子工程本硕;曾任苹果 iPhone/AirPods 高级工程主管;曾从事美国航空航天卫星图像研发。 | 具备顶尖技术功底,跨越硬件工程化、计算机视觉及大规模数据工程。 | 关键;CADDi 自研图纸分析与语义 AI 平台核心架构师。 |
| Ryota Haga | 首席财务官 | 资深企业金融与资本市场高管,负责机构资本运作及全球扩张。 | 主管全球财务运营、美日资本配置与投资者关系。 | 高;对把控资本效率、Series D 资金部署以及筹备后续 IPO 至关重要。 |
| Chris Cope | 工程副总裁 | 资深企业级软件工程管理者,曾带领横跨北美和亚洲的分布式研发团队。 | 负责核心平台工程、云基础设施可靠性及 AI 智能体部署架构。 | 中等;技术执行分散在多位资深工程总监手中。 |
高管档案汇总自 CADDi 官方公司治理披露文件,并经截至 2026 年 9 月的新闻报道证实。
[CO001, CO003, CO004]1.3 资本结构、估值演进与股东治理
CADDi 持续获取全球一线成长期资本支持,通过多轮股权融资跨入独角兽行列。2026 年 9 月,公司就 $114 million Series D 轮融资签署具有约束力的协议,投后估值达 $1.2 billion,较 2025 年 3 月由 Atomico 领投、估值 $470 million 的 Series C 轮追加投资显著跃升。Series D 轮使 CADDi 累计股权融资金额达 $234 million。本轮投资团结构均衡,既有 Atomico、Globis Capital Partners 及 JPS Growth Investment 等全球老股东持续跟投,又引入了 Moore Strategic Ventures、Coreline Ventures、Salesforce Ventures 以及 Toyota Motor 旗下的 Woven Capital 等顶尖战略与财务投资机构。在 CADDi 加速拓展北美市场之际,优质股东背景带来强有力的机构背书与治理监督。 [CO005, CO006, CO007, CO008]
| 投资方 | 投资角色与轮次 | 战略与经济意义 | 尽调焦点与治理权限 |
|---|---|---|---|
| Atomico | 现有领投方(Series C 延期融资领投,参与 Series D) | 欧洲顶尖风投机构,提供全球化扩张支持、欧洲网络资源与持续增长治理支持。 | 拥有董事会席位、同比例优先认购权及后续轮次估值门槛条款。 |
| Globis Capital Partners | 现有领投方 / 早期投资方(跟投 Series D) | 日本顶尖机构 VC,拥有深厚的日本本土工业资源并联通日本邮政银行银团。 | 具备历史董事会影响力,主导本土 IPO 战略协同与大型企业客户关系撬动。 |
| Moore Strategic Ventures | 新进主要机构投资者(Series D) | 全球多资产投资机构,提供成长期资金支持并赋予跨大西洋机构背书。 | 享有保护性条款、知情权及针对后期扩张设立的估值基准考核要求。 |
| Coreline Ventures | 新进主要机构投资者(Series D) | 专注文企软件与工业 AI 基础设施的专业科技投资基金。 | 重点验证制造语义层逻辑;协助导入战略客户资源。 |
| Salesforce Ventures & Woven Capital | 新进战略企业投资者(Series D) | Salesforce 与丰田汽车旗下 CVC 机构,带来广阔的生态分发网络与汽车行业背书。 | 敲定商业合作协议,推动 PLM 与 CRM 集成协同,并在离散汽车制造场景展开试点测试。 |
机构投资方信息摘自 Business Wire 与 SiliconAngle 披露内容;股权结构表持股比例及清算优先权未公开。
[CO007, CO008]1.4 平台演进、产品重构与战略风险
2017 年至 2024 年间,CADDi 曾运营重资产定制零部件采购服务 CADDi Manufacturing,在四个国家布局合作工厂与质检中心。意识到推动制造业转型需要依靠软件基础设施而非实体撮合中介,CADDi 于 2024 年 7 月正式停止全部零部件供应业务,转型为纯粹的 AI 软件平台。2026 年 8 月,公司完成底层平台架构的全面重构,形成四大集成层级:数据层(Data Layer)、语义层(Semantic Layer)、发现与智能体层(Discovery & Agents,含 CADDi Explorer 与 CADDi Agent)以及垂直领域工作流产品(Workflow Products)。尽管技术演进显著,外部风险依然突出。业内质疑指出,CADDi 仍未公开经审计的 ARR 及利润率数据,其 $1.2 billion 估值较大程度依托业务前景叙事。此外,Siemens、SAP、Dassault Systèmes 等传统产品生命周期管理(PLM)巨头具备强大的产品捆绑销售威胁;财富管理领域同名初创企业 Caddi 也带来轻微的品牌辨识干扰。 [CO009, CO010, CO013, CO014, CO015]
CADDi 从定制零件制造采购演进为垂直一体化企业级 AI 平台的发展历程。
[CO001, CO005, CO007, CO009, CO010]1.5 图表
02市场分析
2.1 市场边界、覆盖支出与现状替代方案
CADDi 的目标市场处于离散制造工程数据利用、计算机辅助设计(CAD)归档管理以及采购价值分析 / 价值工程(VAVE)的交汇处。该平台并非核心 3D CAD 设计制图软件的竞争对手,亦不取代企业资源计划(ERP)平台,而是作为一层叠加的智能层运作。它对目前分散在不同企业存储库中的历史 2D 图纸、3D 模型、物料清单(BOM)以及采购记录进行结构化与标准化。覆盖支出包括几何相似性检索、工程图纸光学字符识别(OCR)、自动询价(RFQ)零部件成本核算以及跨业务部门零部件复用发现的软件许可费用。明确排除在外的支出涵盖实体代工采购成本、生成式 3D CAD 建模设计软件以及通用企业文档存储套件。在典型制造工厂中,现状替代方案主要包括人工文件夹检索、缺乏跨工厂几何索引的传统产品数据管理(PDM)或产品生命周期管理(PLM)归档,以及仅储存于资深工程师大脑中的隐性经验。因此,离散制造企业经常因为采购与工程人员无法便捷查阅既有已审批设计,而从不同供应商处以悬殊单价重复采购几乎完全相同的零部件。[CM001, CM002, CM003, CM004]
| 细分市场 | 纳入开支 | 排除开支 | 采购方与买单方 | 战略相关性 |
|---|---|---|---|---|
| 离散机械图纸智能 | AI 图纸录入、几何相似度检索、零件复用索引与 OCR 元数据提取 | 核心 CAD 设计建模许可与生成式 CAD 3D 建模账号 | 采购副总裁、工程主管 | 高:直接替代人工图纸查找与零部件重复建模 |
| 按配置生产(CTO)报价提速 | RFQ 询价图纸解析、自动化成本建模、历史报价检索与供应商比价 | 广泛的 ERP 库存记账许可与财务总账软件 | 报价总监、估算团队主管 | 高:大幅提升定制报价响应速度与中标率 |
| 售后与备件识别 | 历史图纸资产化、跨品牌零部件匹配与备件目录检索 | 实体仓储物流、货运及备件仓储运营 | 全生命周期服务副总裁、售后经理 | 中高:拉动存量设备高毛利售后备件收入 |
| 供应链价值工程(VAVE)与采购优化 | 重复零部件归并分析、价格差异排查与供应商优化 | 实体原材料采购支出与第三方制造交易平台抽佣 | 首席采购官(CPO)、采购寻源总监 | 高:直接降低 10-30% 的零部件采购成本 |
明确 CADDi 软件平台支出对比传统 PLM、CAD 建模及 ERP 预算的划界;支出类别代表企业级软件订阅预算归属。
[CM001, CM002, CM004, CM011]2.2 市场规模测算视角:图纸体量、客群层级与 SaaS ARR
评估潜在市场总额(TAM)需要从多个实证视角审视离散制造业资产,而非依赖单一宽泛的自上而下预测。大型企业里,离散设备制造商保有体量庞大的图纸资产;例如,单一按单工程制造(ETO)包装机械原始设备制造商(OEM)管理的配置选项介于 30,000 至 45,000 种,由此积累的历史工程图纸档案超过 750,000 份。在工业项目中,单厂优化部署通常需要导入 10,000 余份工程图纸以建立零部件基准。从商用软件维度来看,CADDi 于 2022 年由最初的制造业撮合网络模式转型为企业级 SaaS 平台,两年内在日本签约 180 余家企业客户,年经常性收入(ARR)达到 $70 million。客户渗透数据显示,日本前 100 大离散制造企业中已有超 50% 在各业务部门部署该软件。然而,将 AI 图纸数据智能从规模逾 $20B 的更广泛 PLM 及 CAD 软件市场中剥离出来的独立第三方市场测算,仍属待补充的尽调空白。此外,全球供应链波动与关税政策演变正在推高跨国制造中心的软件需求,促使跨国 OEM 必须量化美洲、欧洲与亚洲各工厂的图纸级供应链风险敞口。[CM005, CM006, CM007, CM008, CM013]
| 测算口径与发布方 | 参考年份 | 地理区域 | 预估规模 | 复合年增长率 / 增长趋势 | 测算方法与推导逻辑 | 置信度 | 分析局限性 |
|---|---|---|---|---|---|---|---|
| 企业级图纸智能(自下而上) | 2024 | 日本 | $70M ARR(>180 家客户) | 同比翻倍 | 厂商披露的日本头部 OEM 客户 ARR 及客户数量 | 高 | 仅限日本本土企业级市场,未计入海外收入贡献 |
| 离散制造百强 OEM 渗透率(客户实证) | 2026 | 日本 | >50% 百强制造企业 | 稳步渗透 | 日本最大 100 家制造企业的实证客户渗透率 | 中 | 仅反映企业客户渗透比例,不代表单个客户的实际软件总支出 |
| 企业历史图纸归档量(案例实证) | 2026 | 北美 | 750,000 张图纸(单一 OEM) | 持续增长 | 包装机械 OEM 厂商 nVenia 经审计的客户图纸录入量 | 高 | 仅体现单家企业的资产规模,无法代表全市场 TAM 金额 |
| 全球离散制造 TAM(自上而下口径) | 2026 | 全球 | 缺乏支撑 / 数据缺口 | 预计 AI 软件 CAGR 达 15-20% | 参照离散制造 PLM 与企业搜索软件支出推算 | 低 | 尚无独立分析机构将图纸层 AI 数据利用与广义 PLM/ERP 支出剥离开来 |
综合公司披露、客户实测部署指标及独立研究;因缺乏专注该领域的独立行业分析,全球 TAM 规模仍是尚未解决的尽调缺口。
[CM005, CM006, CM007, CM008]多层规模测算框架,展示离散制造业图纸库规模、大客户导入梯度及已证实的 SaaS ARR。
各层宽度仅表示层级,不代表数值比例。
各测算层级基于公司披露与独立评测中的客户、收入及图纸体量实证梯度,而非无约束的自上而下测算;全球货币化 TAM 仍未公开披露。
[CM001, CM005, CM006, CM007, CM008, CM013]2.3 买方画像细分、预算归属与落地路径
制造数据智能软件的企业采购由离散制造企业内部清晰的多元决策委员会驱动。主要经济买方与预算负责人为采购副总裁、工程主管及首席信息官(CIO),其核心诉求在于遏制利润率下滑并杜绝重复零部件泛滥。日常操作端用户则分化为两大职能:机械设计工程师力求在出图阶段检索并复用既有零部件,报价专员则负责根据历史成本与制造可行性模型评估客户发来的询价单(RFQ)。在按单工程制造(ETO)与按单配置制造(CTO)机械垂直领域,压缩报价周期、抢占高毛利售后备件订单的紧迫竞争压力成为软件落地的触发因素。在精密模具与先进材料制造领域,报价响应速度直接决定客户中标率,促使工程负责人以亚秒级几何相似性查询取代繁琐的多文件夹人工翻找,快速调取既有模具与估算模型。[CM009, CM010, CM011]
| 制造业细分领域 | 目标采购决策者 | 主要使用者 | 买单主体 | 核心工作流 | 预算拥有者 | 采购触发因素 |
|---|---|---|---|---|---|---|
| 按单工程化(ETO)包装机械 | 工程副总裁、全生命周期服务副总裁 | 机械设计工程师、售后专家 | 机械运营事业部 | 检索 75 万张图纸库,复用传感器支架与子装配件 | 全生命周期服务副总裁 | 配置复杂度高(30k-45k 个选项),售后备件检索迟缓 |
| 精密工装与硬质材料 | 首席信息官(CIO)、解决方案副总裁 | 报价工程师、估算专家 | 集团 IT 与商业运营部 | 归集巴塞罗那图纸库,依托历史成本模型为定制 RFQ 报价 | CIO 与商业总监 | 提升 RFQ 响应速度与提高报价中标率的竞争压力 |
| 汽车与重型装备一级供应商 | 首席采购官(CPO)、采购寻源副总裁 | 采购专员、寻源经理 | 采购与战略寻源部 | 录入 10,000+ 张图纸,排查同类零部件供应商报价差异 | CPO / 供应链副总裁 | 供应链关税波动、降本硬性指标与供应商重新谈判 |
| 半导体与多品种工业设备 | 质量副总裁、工程总监 | 质量保证工程师、合规分析师 | 运营与工程事业部 | 交叉比对图纸规格,识别受管制的危险有害物质 | 质量与合规副总裁 | 严格的法规合规要求与资深工程人员退休断层 |
基于客户案例披露与平台部署情况,梳理离散制造垂直领域采购画像、决策层与一线操作人员。
[CM009, CM010, CM011, CM012]2.4 增长驱动因素、实施瓶颈与尽调缺失
工业领域多股宏观趋势正加快产品采用。最突出的是工程人才断层:资深机械工程师与机械师成批退休,几十年积累的隐性知识随之流失,制造商也更容易重复犯设计错误。供应链摩擦加剧、跨境关税变化,还迫使采购负责人依据零件级成本信息迅速重谈供应商条款。另一方面,组织变革和软件集成仍是主要阻力。工程师往往觉得重画零件比翻找分散档案更快;要扭转这一习惯,企业必须持续推动变革,并由厂商客户成功团队协助落地。企业用户也多次指出,平台缺少与内部旧数据库及定制 ERP 工作流之间稳定、双向的 API 集成,迫使客户依赖脆弱的人工桥接。最后,CADDi 在核心日本客群之外可验证的续约率与净收入留存率(NRR)仍未披露,构成平台向西方市场扩张中不可忽视的尽调空白。[CM012, CM013, CM014, CM015]
2.5 图表
03竞争格局
3.1 竞争全景与替代品类别
CADDi 所处的竞争格局涵盖垂直制造智能服务商、传统产品生命周期管理(PLM)巨头、通用云存储工具以及根深蒂固的人工工作流。Krowdbase 等第三方软件聚合平台将 Microsoft SharePoint、Box、ONLYOFFICE Docs 等通用文档管理系统列为 CADDi Drawer 的主要替代品。然而,通用存储平台仅将工程图纸视作未建立索引的二进制文件块,缺乏原生几何计算机视觉、零部件尺寸解析以及供应商成本关联能力。为此,CADDi 累计融资达 $234 million——最近一轮为 2026 年 9 月以 $1.2 billion 估值完成的 $114 million Series D 轮融资——用于在全球制造枢纽拓展专属数据平台。为巩固企业级市场地位,公司正积极优化品牌架构与市场宣传,向采购主管阐明通用办公软件为何无法解决数百万美元规模的供应链效率损耗。[CP001, CP002, CP003, CP007, CP013]
| 竞争对手 / 解决方案 | 类别 | 规模 / 融资额 | 目标客群 | 核心差异化优势 | 局限性 / 权衡取舍 |
|---|---|---|---|---|---|
| CADDi (Drawer / Explorer) | 制造业垂直 AI | 累计融资 $234M(Series D 估值 $1.2B) | 中型到大型离散制造企业 | 多模态图纸相似度检索与自动化 RFQ 报价 | 对接本地旧版 CAD/ERP 数据库需耗费大量集成工作 |
| 企业级 PLM (Siemens / PTC / Dassault) | 传统工程软件套件 | 百亿美元级上市市值 | 全球汽车、航空航天及重型装备巨头 | 覆盖全流程的 CAD 生命周期管控与版本发布工作流 | 部门架构割裂,跨零件自动化成本挖掘能力极弱 |
| 通用 DMS (Microsoft SharePoint / Box) | 通用云存储 | 公有云服务规模(母公司规模 $100B+) | 全行业泛企业知识工作者 | 企业 IT 普及率高,符合 SOC 2 安全要求 | 缺乏原生 CAD 几何解析与二维图纸相似度检索能力 |
| 手工电子表格与内部脚本 | 现状替代方案 | 企业内部日常运营支出(OpEx) | 工厂采购与工程部门 | 软件采购无增量支出,免外部供应商审批 | 检索能力极度受限、人工差错率高且组织知识易流失 |
规模与融资数据根据截至 2026 年 9 月公开披露的创投融资轮次与行业目录信息整理。
[CP001, CP002, CP007, CP008, CP009]制造软件与现有替代方案在图纸智能深度和采购工作流整合度维度的战略定位。
| 条目 | X (1.2–8.8) | Y (1.5–8.2) | 说明 |
|---|---|---|---|
| 1. CADDi (Drawer / Explorer) | 8.8 | 8.2 | 专注 2D 图纸相似度检索,整合采购报价工作流。 |
| 2. 企业级 PLM (Siemens / PTC) | 4.2 | 7.8 | 具备深厚的工程全生命周期管控能力,但跨部门采购情报较为薄弱。 |
| 3. 通用 DMS (Microsoft SharePoint / Box) | 2.1 | 3.5 | 通用企业内容管理工具,缺乏针对 CAD 的计算机视觉能力。 |
| 4. 桌面 CAD 与 PDF 查看器 (ONLYOFFICE / Acrobat) | 3.4 | 2.2 | 文件级查看工具,缺乏多零部件关联索引与历史价格沉淀。 |
| 5. 企业自建定制脚本 | 5.2 | 4.5 | 企业自建临时宏脚本,维护负担重且扩展能力有限。 |
| 6. 现状:人工电子表格 | 1.2 | 1.5 | 传统人工报价方式,各厂区依赖分散的 PDF 文件夹查阅图纸。 |
坐标基于已验证的功能特性与市场覆盖范围,在 0-10 归一化量表上进行有据可查的序数评估。
[CP001, CP003, CP004, CP007, CP012]3.2 技术差异化与图纸智能解析
CADDi 核心架构的差异化优势源于获得专利的相似性检索算法引擎,该引擎将非结构化的历史图纸档案转化为关联企业资产。典型工业客户手头一个项目常涉及 10,000 余份独立工程图纸,历史采购数据往往被困在互不连通的 PDF 文件夹与本地服务器中。尽管 Siemens Teamcenter、PTC Windchill 等传统 PLM 系统能够管理 CAD 版本迭代周期,但通常将技术规格孤立在工程部门内部。CADDi 从 2D 图纸中提取几何轮廓、尺寸公差以及标题栏文本,支持瞬时相似性检索,揭示不同供应商的历史定价偏差。2026 年底,CADDi 推出 CADDi Explorer 与自主智能体 CADDi Agent,推动业务从单一相似性检索拓展至跨部门自动化决策执行。[CP003, CP004, CP005, CP006, CP012]
| 核心能力 / 采购标准 | CADDi | 企业级 PLM 系统 | 通用云端 DMS | 现状手工电子表格 |
|---|---|---|---|---|
| 二维图纸几何相似度检索 | 自研形状与尺寸匹配算法 | 仅限元数据或基础 CAD 模型标签 | 不支持(仅限文本及文件名检索) | 不支持(靠人工肉眼查阅 PDF 图纸库) |
| 采购成本自动估算 | 集成 RFQ 比价与历史报价匹配 | 需定制 ERP 集成并配合人工测算成本 | 不支持(需外接报价工具) | 手工电子表格公式易出现计算差错 |
| 跨部门知识拉通 | 打通设计、采购与质量的统一数据层 | 以工程为中心的信息孤岛,权限设置僵化 | 普通文件库,缺乏行业语义关联 | 分散在员工个人本地电脑中 |
| 自主 AI 工作流智能体 | 2026 年推出行业专用 CADDi Agent | 基于规则的生命周期触发器与脚本自动化 | 通用对话式 Copilot,缺乏 CAD 领域推理能力 | 无(100% 依赖人工) |
功能评估基于核心软件界面展开对比;不支持或未经验证的厂商能力依据公开产品文档判定。
[CP001, CP003, CP004, CP005, CP006]3.3 定价模式、产品组合与采购考量
评估商业软件的取舍,需比较 CADDi 的企业级产品组合与传统产品、替代方案的定价。Stork.AI 的第三方评测指出,自动化后台工具通常为个人用户提供免费增值模式,团队协作则采用定制报价。CADDi Drawer 与 Explorer 按企业年度订阅销售,价格依图纸库体量和用户规模定制。SharePoint 等通用文档工具按席位按月收费,价格透明;CADDi 则需私下协商企业定价和多年维护承诺。条款不透明,采购团队必须提前严格测算 ROI,量化整合供应商和减少重复绘图带来的收益,才能证明这笔支出划算。[CP002, CP010, CP014]
| 解决方案 / 厂商 | 定价模式 | 包含功能 | 折扣策略 / 未知信息 | 采购方考量 |
|---|---|---|---|---|
| CADDi | 定制企业年费订阅 | CADDi Drawer / Explorer 平台、图纸录入与 AI 模块 | 公开口径为团队定制价;多年期折扣阶梯未公开 | 需高管审批预算,并以节省工程工时证明 ROI 合理性 |
| 企业级 PLM (Siemens / PTC) | 实名用户席位费加年度维保费 | 完整 CAD 建模、图纸库归档与生命周期管理 | 多年期企业大单提供大额折扣 | 前期资本开支大,且每年面临供应商深度绑定 |
| 通用 DMS (SharePoint / Box) | 按人头按月 SaaS 订阅 | 云存储、基础文本检索与文档协同 | 标准公开阶梯定价,含企业捆绑折扣 | 采购门槛低,但完全不具备制造图纸智能解析能力 |
商业定价结构根据公开行业目录与披露的打包模式整理;未公开的企业合同约束条款需进一步商业尽调。
[CP002, CP010, CP014]3.4 护城河壁垒、迁移成本与传统巨头反制
CADDi 的长期竞争力取决于数据网络效应,以及与企业价值链的深度集成。Hitachi、Kawasaki、Subaru 和 Tokyo Electron 等工业集团已将 CADDi 嵌入核心寻源业务;数百万历史零部件和供应商报价一旦映射进其语义数据层,迁移成本就会显著上升。整合旧 ERP 数据库与历史图纸需要大量组织投入,客户因此不愿同时使用多个竞品平台。尽管如此,若主流 PLM 巨头开发出原生 AI 相似性检索功能,或客户内部工程团队自建轻量级计算机视觉脚本,CADDi 将面临持续的产品平庸化风险。维持差异化要求 CADDi 持续以实时供应商履约数据充实其语义层,而非单纯依赖静态图纸检索。[CP006, CP009, CP011, CP012]
3.5 图表
04财务表现
4.1 收入来源与商业变现架构
CADDi 采取多层次商业模式变现,将制造业 AI 数据平台的年度软件订阅与实体采购业务收取的交易手续费相结合。核心软件套件以 CADDi Explorer(前身为 CADDi Drawer)及新推出的 CADDi Quote、Design Review 等垂直工作流模块为依托,按经常性企业合同销售。与软件业务并行,CADDi 仍保留部分早期供应链履约业务,通过在其制造网络中撮合定制零部件采购获取利润分成。尽管管理层强调在日本大型制造业客户中渗透迅速,并声称营业总收入实现同比翻倍以上增长,但公司拒绝透露营业收入绝对值、ARR 以及软件与服务业务的收入构成比例。官方公开报价完全基于个案定制且未对外披露,实际成单金额视图纸数据库体量、席位许可数及企业级集成范围协商确定。展示可量化的采购降本成果(例如某工业机械客户零部件成本缩减 30%),依然是 CADDi 企业价值主张的核心。[CI004, CI005, CI006, CI010, CI011]
| 收入流 | 变现机制 | 合同计费单位 | 当前现状与体量 | 收入质量 | 尽调核验诉求 |
|---|---|---|---|---|---|
| CADDi Explorer SaaS | 基于图纸库容量与用户阶梯的年度软件订阅 | 企业年度合同 | 旗舰平台已落地 22 个国家 | 经常性收入质量高;毛利率具备规模效应 | 审查合同 ARR、客户流失率与续约条款 |
| CADDi Agent & Workflows | 模块化增购订阅与按量计费许可 | 单模块年费许可 | 2026 年 8 月推出,覆盖 6 大垂直领域 | 扩张潜力高;早期客户开始采纳 | 核查商机转化率、附加率与折扣力度 |
| 零部件寻源与采购 | 定制制造零部件交易利差与抽成率 | 单笔订单采购金额 | CADDi Manufacturing 旗下的现行业务 | 毛利率较低;背负营运资金与供应链风险敞口 | 评估 GMV 规模、供应商应付款项与减值历史 |
| 驻场部署与客户成功 | 专业入驻导入、定制建模与实施收费 | 工时与物料 / 里程碑合同 | 100+ 名 CS 人员与驻场工程师 | 重服务属性;拉低纯软件毛利率 | 核查计费工时利用率与资本化交付支出 |
CADDi 未披露分部收入具体数据;本表反映已公开商业模式架构与商业化产品线。
[CI005, CI006, CI007, CI010]| 产品 / 组件 | 定价与合同模式 | 牌价与实际成交价机制 | 折扣与信息缺口 | 已披露基准指标 | 尽调要求 |
|---|---|---|---|---|---|
| CADDi Explorer 核心版 | 按图纸库体量计费的企业年度经常性许可 | 企业定制报价;未公开官方牌价 | 折扣与多年期采购承诺及企业集团挂钩 | 已签约日本前 100 大制造商中超 50% | 获取 ACV 分布与平均实际折扣率 |
| 工作流应用(设计审查 / 报价) | 叠加于核心数据平台的分层附加订阅 | 捆绑与独立定价均未公开披露 | 各模块是否设有独立最低企业承诺额尚不确定 | 2026 年发布 6 款垂直领域工作流产品 | 评估现有客户账户的模块附加率 |
| 零部件寻源与制造服务 | 实体零部件采购的成本加成或利差利润率 | 实际利差随供应商报价波动而变化 | 定价面临与日本本土老牌商社的激烈谈判 | 据报客户采购成本降低 30% | 审查各品类零部件的实际抽成率与毛利率 |
| 客户成功与 FDE 支持 | 捆绑企业导入或专项部署合同 | 补贴部署成本以推动企业变革管理 | 未披露 FDE 计费与非计费工时的确切成本分摊 | 聘用超 100 名客户成功人员 | 核查专业服务毛利率与续约附加率 |
官方牌价未公开披露;基准数据反映据报道的商业落地进展与企业定位。
[CI004, CI007, CI011, CI013]从客户图纸接入到 SaaS 订阅增购及实体零件采购毛利的转化路径。
| 节点 / 连接 | 节点 / 起点 | 终点 | 说明 |
|---|---|---|---|
| 节点 1 | 企业图纸库接入 | 客户接入 2D 图纸、3D CAD 文件及历史采购记录 | |
| 节点 2 | 自研 AI 结构化与相似度引擎 | 专利算法把图纸几何特征与历史采购及质量数据相互打通 | |
| 节点 3 | CADDi Explorer 核心 SaaS 订阅 | 提供企业级数据检索与访问权限的年度经常性软件合同 | |
| 节点 4 | 工作流应用与智能体增购 | 针对设计评审(Design Review)、询价(Quote)及特定专业工作流的附加席位授权 | |
| 节点 5 | 零件采购与供应商履约 | 依托 CADDi Manufacturing 合作伙伴网络获取交易抽成与履约服务费 | |
| 节点 6 | 企业综合毛利 | 高毛利软件带来的经常性现金流,会被前线部署工程师(FDE)及交付服务成本抵消 | |
| 连接 1 | 企业图纸库接入 | 自研 AI 结构化与相似度引擎 | |
| 连接 2 | 自研 AI 结构化与相似度引擎 | CADDi Explorer 核心 SaaS 订阅 | |
| 连接 3 | CADDi Explorer 核心 SaaS 订阅 | 工作流应用与智能体增购 | |
| 连接 4 | 自研 AI 结构化与相似度引擎 | 零件采购与供应商履约 | |
| 连接 5 | 工作流应用与智能体增购 | 企业综合毛利 | |
| 连接 6 | 零件采购与供应商履约 | 企业综合毛利 |
展示数据接入、软件经常性合同与实体采购履约之间的结构化转化路径。
[CI005, CI006, CI007, CI011]4.2 GTM 经济模型与服务交付成本负担
CADDi 的商业拓展策略依赖重度介入的企业销售与交付模式,与低摩擦的企业级软件截然不同。鉴于工业制造商面临根深蒂固的组织惯性与庞杂的历史图纸库,推动客户落地必须配合深度的现场变革管理。因此,CADDi 组建了 100 余人的客户成功团队,人数超过纯销售人员,并大力扩招现场交付工程师(forward-deployed engineers)驻场推进系统部署。2026 年 9 月,公司全球总员工数达约 900 人,较 2025 年初的 600 人规模扩张 50%。在美国,CADDi 位于芝加哥的第二总部正积极招募资深 GTM 负责人,基础年薪介于 $100,000 至 $170,000 并附带绩效奖金。这种劳动密集型交付结构带来沉重的持续人力成本负担,拉低了综合毛利率;一旦软件年经常性收入(ARR)增速落后于员工人数扩张,还将引发经营杠杆风险。[CI007, CI008, CI012, CI013]
| 经济指标维度 | 已披露基准 / 代理指标 | 报称数值 | 置信度 | 战略与投资判断关联度 | 尽调核验诉求 |
|---|---|---|---|---|---|
| 年经常性收入(ARR) | 销售额同比翻番(CEO 访谈口径) | 低 | 软件估值倍数的底层分母 | 索取经审计的收入总账及 TTM 经常性收入拆分 | |
| 综合毛利率 | 软件与服务业务占比 | 低 | 服务交付与采购履约拉低整体毛利率 | 要求按软件与服务拆分提供 GAAP 毛利率 | |
| 客户净留存率(NRR) | 企业级增购客户 | 低 | 验证在工程部门之间的自然扩张 | 调取按签约年份划分的队列净留存与总流失分析 | |
| 人力成本结构 | 员工总数 900 人;美国招聘薪资 $100k-$170k | ~900 名员工 | 中 | 表明美日两地每月承担高额固定薪酬开支 | 审计研发、FDE/CS 及高阶销售的薪酬分布 |
缺少可验证公开披露的未公开指标均记为 null,并附上投资人尽调要求。
[CI004, CI007, CI008, CI010, CI012]4.3 资金充裕度、融资依赖与偿债义务
2026 年 9 月完成 $114 million Series D 轮融资后,CADDi 累计获得 $234 million 股权融资。最新一轮融资给予公司 $1.2 billion 投后估值,较 2025 年 3 月的 $470 million 估值实现 2.5 倍倍数扩张。除股权融资外,CADDi 还在 Series C 轮追加轮次中获得来自 Mizuho Bank 及其他主流金融机构的 $30 million 长期债务额度。充裕的资金注入为资产负债表提供充足流动性,重点支持四大既定运营方向:针对工业 CAD 几何特征开发自研 AI 模型、推进平台扩容、加速北美地域扩张以及招募高薪技术人才。尽管在当前人员规模下,现有资金预估可支撑超过 24 个月的运营跑道,但在经常性软件收入足以覆盖日常运营亏损前,公司业务在结构上依然高度依赖外部资本市场或毛利率的大幅改善。[CI001, CI002, CI003, CI009]
4.4 财务研判、收入质量与尽调障碍
从机构尽调视角看,CADDi 是一项进取但财务透明度不足的市场扩张投资。Toyota 旗下 Woven Capital、Salesforce Ventures 等全球顶尖战略与财务投资者愿以 $1.2 billion 估值投资,说明市场认可工业数据基础设施中的垂直 AI。但独立市场观察机构和二级市场追踪平台指出,CADDi 的估值倍数完全建立在未经证实的增长叙事上,缺少经审计的财务基本面支撑。管理层不披露经审计的收入运行率、净收入留存率(NRR)和毛利率,投资者便无法判断公司是在扩大高毛利经常性软件业务,还是靠低毛利的零部件撮合与专业服务支撑收入。要判断 CADDi 的长期资本效率,必须通过保密披露补齐这些私有指标。[CI004, CI010, CI014]
4.5 图表
05产品与技术
5.1 产品矩阵与平台演进
CADDi 围绕垂直一体化软件套件构建商业产品体系,旨在盘活沉淀在制造图纸与物料清单中数十年的工程隐性经验。公司旗舰产品 CADDi Drawer 最初脱胎于支撑其实体零部件制造业务的内部运营软件;彼时工程团队疲于应付零散的图纸服务器与老员工退休带来的人才断层。该系统并非通用文档库,而是构建了一个智能检索环境,对机械图纸、装配体及交叉引用的零件编号建立索引。2026 年 8 月,CADDi 推出下一代平台架构,发布作为 Drawer 升级版本的检索发现引擎 CADDi Explorer,并推出自主人工智能智能体 CADDi Agent,能够在客户设定的运营边界内评估制造数据并执行业务决策。平台还辅以六款针对设计工程、采购及质量管理不同职能定制的垂直领域工作流产品(Workflow Products)。然而,在图纸标准、CAD 格式及用户角色千差万别的异构工业环境中,产品管理与工程团队在标准化软件工作流上面临巨大的复杂度。[CE001, CE002, CE003, CE011, CE012]
| 模块 / 产品线 | 核心用户 | 状态 / 成熟度 | 核心差异化 | 尽调缺口 |
|---|---|---|---|---|
| CADDi Drawer | 设计工程师、采购专员与工厂质量主管 | 全面可用(自 2022 年起作为生产环境核心) | 自研专利 2D 图纸相似度检索,以及老旧图纸扫描件 OCR 识别录入 | 数百万张图纸规模的客户档案库查询延迟等遥测数据 |
| CADDi Explorer | 企业工程负责人与跨职能寻源团队 | 下一代产品上线(2026 年 8 月公布) | 打通图纸几何形状、总成装配树与零件号的统一制造发现引擎 | 生产环境采纳指标与早期 Drawer 部署对比 |
| CADDi Agent | 运营采购、供应商管理人员与制造工程师 | 预览版 / 早期发布(2026 年 8 月公布) | 执行多步骤零件分析与跨系统决策工作流的自主 AI 智能体 | 自主决策错误率与护栏治理机制 |
| 垂直领域工作流产品 | 设计、采购、质量与机械加工各部门专业人员 | 推出 6 款模块化应用(2026 年 8 月公布) | 将自动化决策捕获回填入数据层的上下文感知工作流工具 | 针对专用 MES/LIMS 软件套件与标准 ERP/PLM 连接器的集成深度对比 |
产品线成熟度与差异化特征综合自 CADDi 平台文档与产品公告;运营缺口反映尚未公开的企业遥测数据。
[CE001, CE002, CE003, CE005]5.2 制造工作流、设计复用与成本优化
工业制造企业通过部署 CADDi,化解面向制造的设计(DFM)、采购寻源以及质量保证工作流中的核心数据堵点。在典型的多品种小批量离散制造场景中,查找既有零部件或核验工程技术规范,需要跨越孤立的企业资源计划(ERP)数据库、共享网络文件夹与实体图纸档案柜进行繁琐的人工核对。CADDi 借助能够从传统 2D 蓝图、扫描 PDF 及手写批注中提取文本与标注的光学字符识别(OCR)技术,结合基于形状检索历史零部件的专利几何相似性算法,消除了这一操作痛点。通过在历史档案中检索出完全相同或功能可替代的零部件,平台协助工程师前置开展可制造性评审,避免重复开模。在采购端,买方利用几何聚类识别相似零部件的成本差异,赋能价值分析 / 价值工程(VAVE)与供应商商务谈判。尽管如此,工作流效率仍严格受制于原始数据质量;若导入的图纸库存在严重的扫描褪色、命名标注不一或尺寸标注错误,这些缺陷将直接传导至下游推荐结果中。[CE003, CE004, CE005, CE009, CE010, CE013]
| 用户任务 | 现状工作流 | 公司解决方案 | 可量化收益 | 局限性 |
|---|---|---|---|---|
| 历史零件号与图纸检索 | 在孤立的 ERP、共享网盘与纸质文件柜中手工查询 | CADDi Drawer 统一检索,对图纸、总成及交叉引用的零件号建立索引 | 用单次查询检索 100 年归档档案,取代依赖个人经验的多步骤检索 | 受制于初始 OCR 扫描质量与元数据索引规范度 |
| 设计复用与可制造性设计(DFM)分析 | 在无法看清公司现有零部件库的情况下,从零重复设计定制零件 | 专利几何相似度检索,调出历史设计及关联的缺陷与质量报告 | 消除重复新建零件,在开模投入前加快设计评审周期 | 需要 2D 图纸标准化,且无法替代完整的 3D 有限元仿真 |
| VAVE 与采购成本对标 | 在缺乏同类零件历史价格差异背景的情况下,进行临时供应商谈判 | 自动相似度分组,将零件几何形状与历史采购订单及供应商报价关联 | 发现同类零部件的价格差异,推动采购量归集与目标成本谈判 | 需要 ERP 采购订单持续双向同步以保持价格实时更新 |
| 受控物质与合规审计 | 人工审查文档,追踪供应商材质证明与图纸批注 | 自动全文 OCR 提取,解析工程附注与材料规格标注 | 据报排查零件库受控物质使用情况所需时间减少 90%+ | 客户自述案例指标;边缘手写附注仍需人工核验 |
工作流对比来自 CADDi 案例研究与工程招聘启事;据报百分比收益反映具体客户的营销宣传口径,并非经审计的跨客户平均值。
[CE003, CE004, CE005, CE009, CE010]5.3 四层技术架构、核心技术栈与部署模式
CADDi 平台依托四大集成功能层级运行:数据摄入与分析的数据层(Data Layer)、上下文与关系的语义层(Semantic Layer)、发现与智能体的智能层(Discovery & Agents Layer)以及执行层面的工作流层(Workflow Layer)。底层工程实现融合现代 Web 应用框架与系统级编程语言,解决方案架构师与系统工程师采用 Python、TypeScript、Rust 配合 SQL 编排编写生产流水线。基础设施以容器化 Docker 工作负载形态部署,依托持续集成与交付流水线托管于 Amazon Web Services、Google Cloud Platform、Microsoft Azure 等主流公有云厂商。为服务国防工业领域企业客户,CADDi 执行严格的安全控制标准,美国团队解决方案交付严格遵循《国际武器贸易条例》(ITAR)人员资质审查规定。工程团队在人才招聘中引入结构化技术架构考核,评估候选人在数据库架构设计、云基础设施与数据管道高可用性等维度的权衡决策能力,以支撑高吞吐量工业级分析。[CE001, CE004, CE005, CE006, CE007, CE008, CE014]
| 架构分层 / 组件 | 角色与功能 | 核心依赖 | 技术与运营风险 |
|---|---|---|---|
| 数据摄取与光学识别层 | 摄取存量 2D 蓝图、PDF 扫描件与手写工程附注 | 专有 OCR 模型、文档解析引擎与安全图像预处理流水线 | 若纸质历史图纸扫描件老化、字迹不清或缺失关键视角,将面临垃圾进垃圾出风险 |
| 语义关系与图谱层 | 梳理图纸与总成、BOM、供应商规格书及零件号的上下文关联 | 实体抽取、元数据标准化器与企业 ERP/PLM/CAD 关系型连接器 | 外部 ERP 交易记录与 CADDi 图谱模型之间的模式漂移和同步延迟 |
| 向量索引与相似度检索引擎 | 提取几何轮廓与形状向量嵌入,支持亚秒级图纸相似度检索 | 专有专利相似度检索算法与高维向量索引结构 | 随着数千万件工程资产规模膨胀,查询性能下滑及索引内存占用飙升 |
| 应用集成与云端交付 | 交付多租户 Web 应用、API 集成与容器化部署基础设施 | Python、TypeScript 与 Rust 后端、Docker 容器、AWS/GCP/Azure 云服务及 ITAR 合规 | 企业部署周期复杂、定制网络防火墙审批繁琐,以及 30%+ 现场出差要求 |
运行架构组件汇编自 CADDi 技术招聘、工程师岗位职责说明与架构白皮书;独立的延迟与错误率基准数据未公开披露。
[CE001, CE004, CE005, CE006, CE007, CE008, CE013]层级技术栈,展示 CADDi 从底层企业云基础设施到顶层特定领域工作流应用的 6 层架构。
- 工作流应用产品
- 面向设计复用、采购优化、制造工程与质量管理的特定领域模块化应用。
- 检索发现与自主智能体层
- CADDi Explorer 搜索引擎与自主式 CADDi Agent,执行多步骤分析与运营决策工作流。
- 语义关系层
- 上下文知识图谱,将图纸几何结构与物料清单(BOM)、交叉对照零件号以及 ERP/PLM 元数据打通。
- 几何特征与向量相似度引擎
- 专利 2D 形状提取算法与高维向量检索,对零部件轮廓与视觉特征建立索引。
- 数据接入与 OCR 处理层
- 文档解析流水线,从 2D PDF 与纸质蓝图中提取文本、标题栏、公差及手写注释。
- 云基础设施与企业级集成层
- 用 Rust/Python/TypeScript 构建容器化 Docker 微服务,部署在 AWS/Azure/GCP,并落实 ITAR 合规管控。
架构体系综合自 CADDi 平台技术规范、Series D 轮技术发布及工程招聘岗位架构。
[CE001, CE002, CE004, CE006, CE007, CE008]5.4 技术差异化、合规管控与尽调风险
CADDi 的主要技术壁垒建立在专为机械工程图纸定制的领域专用 AI 模型之上;通用大语言模型与通用搜索引擎难以准确解析正投影视图、几何尺寸及形位公差标注。公司拥有专利的相似性检索技术让工程师仅凭草图或 2D 轮廓即可锁定既有零部件,无需依赖规整的 3D CAD 模型或预先打好的元数据标签。在监管与合规端,CADDi 遵循 ITAR 人员准入审查协议,为服务北美大型 OEM 构筑了有力的采购资质准入门槛。然而,技术尽调显示出显著的运营风险:针对低质量图纸的 OCR 错误率、数百万级图纸体量下的向量索引查询延迟,以及与本地部署旧 ERP 数据库的跨系统同步延迟,平台均未见公开验证的第三方审计基准。此外,企业级部署面临繁重的上线摩擦,客户需在专属解决方案架构师协助下完成深度数据映射,方能实现自动化工作流收益。[CE005, CE008, CE010, CE012, CE013]
5.5 图表
06客群分析
6.1 客群画像细分与目标企业定位
CADDi 重点服务复杂且文档密集的离散制造企业,此类客户在数十年的运营中积累了大量彼此割裂的工程设计、采购历史与质量记录。核心买方画像涵盖一级汽车零部件供应商(Tier-1)、工业设备原始设备制造商(OEM)以及高精密度外协加工厂的设计工程副总裁、采购总监及质量控制负责人。此类企业通常在 SAP、Infor LX 等 ERP 系统以及传统网络文件服务器中管理着数万至逾百万份工程图纸、CAD 文件与历史物料清单。CADDi 通过两大互补渠道向客群创造价值:一是核心软件应用 CADDi Drawer(正演变为 CADDi Explorer 及垂直工作流工具);二是保障零部件质量、交期与价格的轻资产供应链采购服务 CADDi Manufacturing。尽管更宽泛的企业自动化市场还包含面向后台流程挖掘的工作流方案,但 CADDi 的核心竞争壁垒源于其对重工业制造资产专属的几何与语义解析能力。[CU001, CU002, CU010, CU011, CU014]
| 客户群组 | 目标买家与用户 | 主要使用场景 | 观察到的部署规模 | 收入与战略价值 | 关键尽调缺口 |
|---|---|---|---|---|---|
| Tier-1 汽车零部件供应商 | 设计工程副总裁 / 质量控制经理 | 图纸统一检索、历史 QA 归因分析、供应商零部件报价 | 单厂 30 至 85 个活跃账户;录入 140,000+ 份图纸与 QA 表单 | 战略价值高;具备多年期企业授权潜力与全厂推广空间 | 跨海外基地的多厂推广节奏及全球合同条款未经核实 |
| 重型装备与工业 OEM | 质量主管 / 机械工程总监 | 装配反查检索、老旧 ERP 交叉比对、杜绝重复建件 | 数千份图纸、复杂 BOM 总成及多层级子部件目录 | 打通设计与采购带来高 ARR 潜力;削减重复采购订单支出 | 合同续约期及相较内部工程数据库的软件锁定效应 |
| 精密金属与塑料加工商 | 销售工程 / 模具估价专员 | 历史 RFQ 报价匹配、形状相似度检索、利润率精准核算 | 录入超 1,000,000 份文档;75% 的 RFQ 业务量经由平台处理 | 关键运营价值;将报价周期从 30 分钟压缩至 2 分钟 | 依赖客户报价单量及周期性制造需求 |
| 企业工作流自动化(trycaddi) | 运营负责人 / 业务管理总监 | 只读桌面工具流程挖掘与自动化工作流闭环 | 已连接 10 至 150+ 款工具(Outlook、Clio、Salesforce、DocuSign) | 从免费增值到 $6,500/月分层变现;自主快速导入 | 系独立于 CADDi 核心制造数据平台之外的企业实体,有待进一步厘清 |
综合自 CADDi 已发布案例、技术通告与官方定价文档;规模指标仅反映单厂客户披露口径。
[CU001, CU003, CU005, CU007, CU010, CU014]6.2 使用渗透轨迹与日常黏性跃升
CADDi 数据平台的落地通常始于具体的部门痛点——尤其是人工检索历史图纸、核对零部件版本以及针对客户询价单(RFQ)比对历史报价所损耗的大量工程工时。一旦完成部署,随着日常工时节约在相邻车间流程中逐步显现,软件使用呈现迅速的自下而上渗透。在一级汽车零部件供应商 Neaton Auto Products Manufacturing,最初部署时有 30 名用户;质检与采购团队将平台融入日常工作后,数周内活跃账户自然增长 183%,达到 85 个。同样,在精密金属与塑料加工制造企业中,模具估价专员已在约 75% 的 RFQ 业务中采用 CADDi。图纸检索时间从原先的 30 余分钟普遍缩减至一到两分钟,效率提升超过 70% 至 96%。这种业务黏性促使 CADDi 从一款试验性检索工具升格为厂区不可或缺的关键基础设施,在加快投标响应速度的同时避免高昂的零部件重复开发成本。[CU002, CU003, CU004, CU007, CU008]
| 采纳指标 | 报称数值 | 统计日期 | 来源出处 | 置信水平 | 战略意涵 | 缺失的分母数据 |
|---|---|---|---|---|---|---|
| 全球客群版图 | 覆盖 22 个国家企业 | 2026-09 | Simplify / CADDi 披露 | 中 | 证实业务已突破日本本土核心基本盘,走向国际市场 | 未披露活跃付费账户总数与试点部署对比 |
| Neaton 用户扩张 | 活跃账户增至 85 个(较 30 个增长 183%) | 2026-05 | CADDi 官方案例研究 | 中 | 展示席位从设计向质控延伸的自下而上自然扩张 | 俄亥俄州伊顿工厂潜在工程师总编制 |
| RFQ 估价使用率 | 占 RFQ 总量 75% | 2026-01 | Business Wire / CADDi 新闻稿 | 中 | 证明模具估价人员在日常工作中高度依赖 CADDi Drawer | 月度报价总次数及挂钩确认的制造收入 |
| 图纸检索耗时缩减 | 缩短 40% 至 70%(降至 1.5 分钟) | 2026-05 | Neaton 案例研究 | 中 | 大幅节省日常工程工时,加快根因调查排查 | 整个工程部门累计节省工时总数 |
| 报价检索周转耗时 | 从 30 分钟缩短至 1 至 2 分钟(降幅 96%) | 2026-01 | 高精度制造厂商发布通告 | 中 | 大幅压缩投标响应周期,拉升中标率与估价处理能力 | 基准估价人员编制与每日提交的竞标报价单量 |
所有运营表现数据均来自公司披露的客户案例;群组留存率与客户总量未公开披露。
[CU002, CU003, CU004, CU007, CU008]从初期图纸检索痛点到多部门工作流集成的企业客户落地生命周期。
各阶段与触点综合自 Neaton、Amerequip 及工业精密制造企业的公开客户落地案例。
[CU003, CU004, CU005, CU006, CU007, CU009]6.3 标杆客户实证与落地运营成效
公开记载的实际生产部署为 CADDi 在北美与日本制造厂区的运营可行性提供了坚实佐证。美国 OEM 厂商 Amerequip Corporation 为 John Deere、Caterpillar 及 Oshkosh Corporation 等全球客户设计并制造定制设备与液压系统;其工程与质量负责人借助 CADDi Drawer 打通图纸、客户技术规范与 ERP 记录。Amerequip 团队告别了繁琐的共享网络磁盘翻找,利用关键词与相似性检索迅速识别库存中已有的完全相同零部件,有效规避不必要的采购订单与重复开模成本。在俄亥俄州伊顿的 Neaton Auto Products Manufacturing,工程师集中归档了 140,000 余份工程与质量技术文件,图纸查阅工时缩减 40% 至 70%,质量保证根因调查提速 50%。在日本本土,CADDi 服务于 Hitachi、Subaru 及 Kawasaki 等老牌工业企业,据报道在日本前 100 大制造企业中的渗透率已过半。[CU003, CU004, CU005, CU006, CU007, CU008]
| 客户机构 | 行业细分 | 部署与使用场景 | 生产环境 vs 试点验证 | 实测运营成果 | 尽调局限 |
|---|---|---|---|---|---|
| Neaton Auto Products Manufacturing (NAPM) | Tier-1 汽车内外饰件 | 汇集 PLM/ERP 中 140,000 份文档,用于图纸检索与 QA 根因分析 | 生产环境 | 85 名活跃用户(增长 +183%);图纸检索耗时减少 40-70%;QA 检索耗时缩减 50% | 仅为单厂部署数据;跨厂企业级推广尚待确认 |
| Amerequip Corporation | 定制设备与液压制造(OEM) | CADDi Drawer 与 LX ERP 集成,解决客户零件号不匹配与 BOM 反查问题 | 生产环境 | 省去在多个网盘间来回查找;避免重复新建零件与冗余采购订单 | 未量化成本节约幅度;持续依赖客户工程团队配合 |
| 高精度金属与塑料制造厂商 | 精密工业与功能部件 | 模具估价与销售工程 RFQ 图纸相似度检索 | 生产环境 | 报价检索耗时从 30 min 缩减至 1-2 min;75% 的 RFQ 单量由 CADDi 处理 | 公告中隐去客户名称;定价结构与合同金额未公开 |
| 日本工业巨头(Hitachi、Subaru、Kawasaki) | 重工业、汽车与交通运输 | 部署 CADDi Drawer/Explorer,优化全制造价值链供应链 | 生产环境 | 据称在日本前 100 大制造企业中已有过半采纳 | 未披露具体部署深度、合同金额及各产品模块拆分 |
列举内容代表已发布的企客案例以及经新闻稿核实的客户佐证;所覆盖样本仅选自公开披露材料。
[CU001, CU002, CU003, CU004, CU005, CU006, CU007, CU008, CU012]6.4 客户留存、增购机制与集中度风险
尽管客户定性反馈与落地热情显著,评估 CADDi 的商业持续性仍暴露出关键尽调盲区。CADDi 目前未公开经审计的总收入留存率(GRR)、净收入留存率(NRR)以及年度客户流失率,导致投资者与企业合作伙伴只能依赖营业总收入增长叙事,而无法获取经审计的同客群留存瀑布图(cohort waterfall)。为稳固客户留存,CADDi 在北美招募专属客户成功主管,专门负责在涉及多方利益相关方的复杂工业销售周期中建立售后服务手册并优化 NRR 目标。此外,CADDi 面临来自传统软件巨头的长期续约威胁。主要的 CAD、PLM 与 ERP 供应商——尤其是 Siemens、SAP 及 Dassault Systèmes——在全球制造业中保有深厚的客户合作关系,完全可能在 2027 年前将 AI 图纸检索功能捆绑进既有企业合同。一旦北美市场的“拓客与增购”(land-and-expand)策略遇阻,客户数据团队可能会尝试利用企业内部数据湖仓自建平替,而非续订该软件平台。[CU009, CU012, CU013]
6.5 图表
07风险评估
7.1 风险分类与综合研判热力图
CADDi 从技术型零部件采购撮合商转向企业级 AI 数据平台,改变了其运营与战略风险。2026 年 9 月以 $1.2 billion 估值完成 $114 million Series D 轮融资后,公司拓展北美市场也面临更高预期和执行压力。主要风险有四类:老牌 PLM 套件带来的结构性竞争、包括美国国防出口管制在内的跨境合规要求、处理历史工程图纸时的技术精度,以及独立软件留存指标不透明带来的治理风险。CADDi 无法像部分垂直软件同行那样,受益于强制监管申报要求带来的保护;客户留存因此取决于 CADDi 能否在传统 CAD 巨头推出捆绑 AI 功能前,持续积累专有图纸智能并深度嵌入企业工作流。[CR002, CR003, CR005, CR006, CR014]
风险矩阵,梳理 CADDi 主要企业风险类别的固有发生概率、固有严重程度、缓解措施成熟度及残余风险。
各项评估反映综合自独立 SaaS 市场分析、行业评论及公司披露的定性研判。
[CR002, CR003, CR005, CR008, CR009, CR014]7.2 监管合规、ITAR 出口管制与法律风险
CADDi 加速渗透一级汽车零部件、重型机械与航空航天制造业之际,监管合规成为左右销售推进周期的关键因素。2026 年 8 月,CADDi 在其 CADDi Drawer 软件平台中上线符合 ITAR 标准的合规功能,以支持航空航天与国防承包商处理受美国国务院国防贸易管制局(DDTC)监管的技术数据。该项业务必须实行严格的美国本土数据驻留、加密权限访问控制以及物理隔离的云环境,防范外籍人员违规访问。此外,CADDi 的跨国运营版图覆盖日本、美国、越南与泰国,客户遍及 20 多个国家,令其面临互有差异的跨境数据传输法规、用工政策及出口管制约束。国防采购合规疏漏或涉密国防 CAD 图纸泄露,可能引发巨额法定罚款、取消承包商资格以及灾难性的商誉毁损。[CR001, CR015]
| 合规规定 / 监管要求 | 司法管辖区 | 状态 | 发生概率 | 严重程度 | 缓解策略 | 残留风险敞口 | 尽调核验路径 |
|---|---|---|---|---|---|---|---|
| ITAR / DDTC 出口管制 | 美国 | 平台现已支持 | 中 | 极高 | CADDi Drawer 部署执行严格的美国本土数据驻留、分区隔离访问及符合 DDTC 规范的基础设施 | 中等;国防承包商的操作失误可能导致受控 CAD 文件外泄 | 核查第三方安全审计报告与 DDTC 合规认证 |
| 跨境数据传输与 IP 保护 | 美国、日本、欧盟、越南 | 持续合规中 | 中 | 高 | 跨国子公司之间实行客户独占云租户与加密边界隔离 | 中等;各地区隐私法规与云主权指令存在差异 | 核查多区域 AWS / 云数据驻留架构与数据处理协议(DPA)合同 |
| AI 治理与自主智能体责任 | 美国及全球 | 监管框架处于成形阶段 | 低 | 高 | CADDi Agent 严格在公司定义的语义规则与人工在环审批下运行 | 低;软件自主决策的新兴监管标准尚在酝酿 | 调阅 CADDi Agent 审计日志、安全护栏与责任免责声明 |
| 历史商业合同负债 | 日本 | 已终止运营(2024 年 7 月) | 低 | 中 | 彻底关停实体零部件制造与采购撮合市场业务 | 极低;2024 年前交易所遗留的质保或供应商争议尾部风险 | 索要撮合业务关停的正式法律责任解除书与赔偿免责附表 |
风险评估基于美国国务院 DDTC ITAR 管制要求、全球贸易法规及上市公司披露材料;各行按严重程度排序。
[CR001, CR015]7.3 系统运行可靠性、模型准确度与团队扩张瓶颈
CADDi 的技术价值依托自研 AI 模型、光学字符识别(OCR)与几何相似性检索算法来录入并结构化工程图纸。然而,离散制造客户往往沿用数十年前建立的旧 IT 系统与工程归档库,充斥着非标制图符号、手写批注与低清晰度扫描件。尺寸提取偏差或形位公差分类错误可能向下游传导,导致零部件采购、供应商竞标及装配线作业出错,引发客户索赔纠纷。在日常运营中,CADDi 面临团队地域分布不均带来的组织扩张瓶颈:根据公开画像统计,全球团队约 77% 仍驻扎在日本,而核心增长重心已转向美国。在技术人才争夺激烈的背景下,芝加哥企业级销售团队的组建亟需兼通双语的复合型增长管理人才,以协调日本技术工程团队与美国制造高管间复杂的 GTM 落地协同。[CR004, CR008, CR009, CR010, CR012]
| 失效模式 | 发生概率 | 严重程度 | 缓解措施成熟度 | 残余敞口 | 未决缺口 |
|---|---|---|---|---|---|
| 图纸 OCR 与几何相似度误分类 | 中 | 高 | 成熟 | 中等;边缘工件公差可能误导采购 | 手写历史图纸的 OCR 准确率实证基准数据 |
| 传统 IT 技术栈与 CAD/PLM 接入阻力 | 高 | 高 | 发展中 | 高;30-40 年历史的老旧系统延缓企业价值实现周期 | 客户平均上线周期与专业服务工时 |
| 跨国人才与双语管理层瓶颈 | 高 | 中 | 发展中 | 中等;双总部协同放慢决策速度 | 美国团队招聘进度与日本工程团队协同 |
| 专有数据接入与知识产权泄露风险 | 低 | 极高 | 成熟 | 低;客户数据按企业严格隔离 | SOC 2 Type II 报告与第三方渗透测试结果 |
基于工程平台公开信息、GTM 架构分析与客户落地案例评估;各行按严重程度排序。
[CR004, CR008, CR009, CR010, CR012]7.4 传统 PLM 巨头捆绑策略、生态依赖与网络效应
CADDi 的长期战略壁垒极易受到 Siemens、SAP、Dassault Systèmes 以及 PTC 等全球工程软件巨头反制策略的冲击。这些巨头牢牢掌控着企业核心的产品生命周期管理(PLM)与 CAD 系统,维系着数十年的主服务协议与深厚的采购关系。PLM 厂商有能力在 2027 年前将原生 AI 图纸检索和相似性匹配直接打包进现有软件许可,从而削弱 CADDi 单点解决方案的价值主张。不同于具备多边网络效应的企业级平台,鉴于严格的知识产权隔离墙,CADDi 运行在封闭的企业专有数据集之上,无法实现跨企业图纸互通。因此,客户迁移成本并非在第一天自然形成,而是随着使用时长逐步沉淀。尽管 CADDi 将自身定位为协同共享的智能增强层而非直接替代 PLM 与 ERP 底座,但在预算收紧的背景下,企业 IT 部门往往更倾向于选用 PLM 原生扩展模块,而非单独采购第三方软件订阅。[CR002, CR003, CR004, CR013]
| 依赖类别 | 合作方与生态系统 | 业务职能 | 集中度 | 潜在失效场景 | 严重程度 | 缓解措施 | 残余敞口 |
|---|---|---|---|---|---|---|---|
| 核心企业级 PLM/CAD 生态系统 | Siemens, SAP, Dassault Systèmes, PTC | 托管核心工程 CAD 数据与图纸 | 极高 | 老牌厂商将原生 AI 相似度检索打包进核心 PLM 授权 | 高 | 定位为互补的语义层,并拓展至 CADDi Quote/Agent | 高;缺乏合规锁定壁垒或跨企业网络效应 |
| 大客户集中度 | 日本汽车与机械主机厂(Subaru、Denso、Hitachi) | 支撑早期收入与图纸处理规模 | 高 | 头部客户流失或预算冻结导致增购收入缩减 | 高 | 依托芝加哥销售中心积极拓展美国市场以分散风险 | 中等;收入仍高度集中于日本前 100 大制造商 |
| 商业债务与银行银团 | Mizuho Bank 与银团参贷机构 | 提供 $30M 长期债权融资 | 中等 | 增速放缓时,偿债契约限制研发资金调配 | 中 | $114M Series D 轮股权融资增强流动性缓冲 | 低;股权融资带来的现金储备提供充足的跑道缓冲 |
| 云基础设施与 AI 算力供应商 | 超大规模云厂商(AWS / Google Cloud / Azure) | 托管多区域数据层与 AI 推理模型 | 高 | API 涨价、云服务宕机或区域服务中断 | 中 | 多区域云冗余与容器化平台架构 | 低;通用算力易于跨云厂商抽象解耦 |
依赖体系分类评估企业软件生态伙伴、基础设施提供商与资本银团风险敞口;各行按严重程度排序。
[CR002, CR003, CR007, CR013]7.5 财务状况、债务承诺与监测触发项
CADDi 为搭建软件底层耗资巨大。2025 年 3 月 Series C 延期融资时,公司除拿下 Mizuho 牵头银团提供的 $30 million 长期银行授信外,股权融资累计筹集 $234 million。2026 年 9 月,CADDi 以 $1.2 billion 估值完成 $114 million Series D 轮融资,但未公开披露独立软件的年经常性收入(ARR)、总流失率或净留存率(NRR)。核心指标不透明迫使投资者依仗收入扩张叙事,而非经审计的 SaaS 单位经济模型。机构投资者判定逻辑破损的核心触发点包括:北美企业客户续约未达日本本土历史净留存水平;传统 PLM 巨头激进捆绑使得年度新客户增速跌破 40%;企业软件现金流无法覆盖美国市场高昂的商业化烧钱开支,诱发偿债压力。[CR005, CR006, CR007, CR011, CR014]
7.6 图表与附录
08估值
8.1 投资逻辑与建议框架
工业制造与企业级人工智能交汇处,CADDi 展现出吸引力与复杂性并存的投资特征。核心投资逻辑在于,CADDi 证实自身有能力突破管理层所谓的「物理瓶颈」,也就是生成式软件能力与工厂车间落地之间的长期鸿沟。通用基础模型无法稳定解析复杂二维工程图纸或三维 CAD 几何体,CADDi 则自研行业专用视觉模型,搭配通用 LLM 处理文本图纸。这种技术差异化得到日本本土企业级渗透率的强力支撑,日本前 100 大制造企业超半数采用该平台。然而,估值纪律要求保持审慎。公司以 $1.2 billion 估值拿下 $114 million Series D 融资,在缺乏 ARR 规模或软件毛利率公开验证的前提下,估值已达顶级独角兽水平。因此,机构投资者宜采取 Track / Research-More 观察立场,暂缓正式出资,等待单位经济模型与欧美市场商业拓展成果得到独立证实。[CV016, CV002, CV003, CV007, CV013, CV014, CV015]
| 维度 | 立场 | 评估结论 | 决策含义 |
|---|---|---|---|
| 投资建议 | 观察 / 继续研究 | 围绕实体制造数据建立的垂直 AI 壁垒明确,但收入与利润率未披露,入场仍需守住纪律。 | 在商业化 ARR 规模、客户集中度及单客留存得到证实前,暂不投入初始资金。 |
| 置信度 | 中 | 已由顶级银团完成估值 $1.2B 的 Series D 轮融资,但基本面损益表指标仍未公开。 | 在正式作出投资判断前,优先获取经审计的利润表与毛利率结构。 |
| 风险等级 | 高 | 变革管理开销庞大(100+ 名客户成功人员),且依赖传统制造业较长的导入周期。 | 重点跟踪客户增购效率、前线部署工程团队(FDE)占比以及国际市场拓展势头。 |
| 估值立场 | 偏高 | 收入未披露,估值已达 $1.2B、约为累计融资总额的 5.1 倍,几乎没有为执行失误留出安全边际。 | 争取折价通过老股交易入场,或待 Series D 里程碑兑现后再考虑追加投资。 |
基于 CADDi 2026 年 9 月 Series D 融资条款及已披露运营指标的综合研判。
[CV016, CV002, CV003, CV007]决策链路,将企业级客户验证、融资规模、组织变革开销及估值透明度缺失连接至“观察”评级。
| 节点 / 连接 | 节点 / 起点 | 终点 | 说明 |
|---|---|---|---|
| 节点 1 | 企业壁垒与客户验证 [node-proof] | 日本前 100 大制造企业中超过 50% 部署;自研图纸 AI | |
| 节点 2 | 资本储备与业绩增长 [node-scale] | $114M Series D 轮融资,估值 $1.2B;累计融资 $234M;销售额同比翻倍 | |
| 节点 3 | 结构性执行风险 [node-risk] | 客户成功团队开销高昂(100+ 人);海外拓展与组织变革阻力 | |
| 节点 4 | 估值与信息透明度缺口 [node-valuation] | 估值 $1.2B 且未披露 ARR 和利润率;Dealroom 基准排在第 38 百分位 | |
| 节点 5 | 估值立场:偏高 [node-stance] | 溢价倍数默认未来执行近乎完美;入场安全边际狭窄 | |
| 节点 6 | 投资建议:观察 / 继续研究 [node-recommendation] | 资金投入前须确认经审计损益表、客户队列留存率以及美国拓展成效 | |
| 连接 1 | 企业壁垒与客户验证 [node-proof] | 估值立场:偏高 [node-stance] | 技术验证 |
| 连接 2 | 资本储备与业绩增长 [node-scale] | 估值立场:偏高 [node-stance] | 资金背书 |
| 连接 3 | 结构性执行风险 [node-risk] | 估值立场:偏高 [node-stance] | 利润率拖累 |
| 连接 4 | 估值与信息透明度缺口 [node-valuation] | 估值立场:偏高 [node-stance] | 估值溢价 |
| 连接 5 | 估值立场:偏高 [node-stance] | 投资建议:观察 / 继续研究 [node-recommendation] | 尽调前置条件 |
逻辑梳理,阐释自研技术防御壁垒与入场估值纪律之间的权衡。
[CV016, CV003, CV007, CV013, CV014, CV015]8.2 核心支柱评估:看多与看空逻辑
评估 CADDi,需要权衡强劲的运营顺风与重大的执行风险。看多逻辑在于,公司从 CADDi Drawer 的采购去重工具起步,逐步拓展为涵盖 CADDi Explorer、CADDi Agent 及 6 款工作流产品的全套制造 AI 数据平台(Manufacturing AI Data Platform)。Atomico、Woven Capital(丰田)、Coreline Ventures 以及 Salesforce Ventures 等头部战略与财务机构注资,赋予其极高的商业信誉与生态资源。反观空方逻辑,风险主要集中在组织架构与市场竞争。CADDi 拥有 100 多名客户成功人员,人数超过销售团队,专职驻厂推行深入的变革管理。极高的交付服务密度印证了改变制造企业数十年固有习惯的阻力,也可能把高毛利纯软件业务拉低成科技赋能服务业模式。不仅如此,通用基础模型厂商或 Autodesk、Siemens、Dassault Systèmes 等传统 CAD/PLM 巨头若推出具备几何推理能力的竞品智能体,也可能逐步蚕食 CADDi 的技术溢价。[CV004, CV005, CV006, CV008]
| 核心支柱 | 投资逻辑 | 反向逻辑 | 逻辑破局与观点转变触发条件 |
|---|---|---|---|
| 自研数据护城河 | 自研 AI 模型能够解析通用大语言模型无法处理的 2D 工程图纸与 3D CAD 文件,构建起技术防御壁垒。 | 通用基础模型可能具备原生多模态几何推理能力,或传统 CAD 软件厂商可能缩小技术差距。 | 基础大语言模型无需垂直领域微调即可在复杂多视图工程图纸解析上达到同等水平。 |
| 企业级客户拓展 | 日本前 100 大制造商中超过半数采用该平台,形成企业级背书与供应链网络效应。 | 出海日本以外市场面临根深蒂固的老旧 ERP/PLM 系统集成困难、供应商标准繁杂以及本地变革管理阻力。 | 北美市场收入未能起量,或销售周期拉长超过 12 个月。 |
| 产品平台化 | 从单点工具(CADDi Drawer/Explorer)迈向多工作流数据平台及自主式 CADDi Agent,拓宽了 ACV 与 TAM 空间。 | 客户成功团队规模庞大(100+ 名 CS 人员,人数超过销售),反映服务交付重度,可能拖累 SaaS 利润率。 | 因前线部署工程师(FDE)与定制集成成本导致毛利率跌破 65%。 |
| 资本结构与投资阵容 | 获顶级全球投资者支持,包括 Atomico、Woven Capital(Toyota)、Coreline 与 Salesforce Ventures,累计融资达 $234M。 | 在未披露收入的情况下,估值于 18 个月内翻倍至 $1.2B,在风投资本收紧周期下面临估值倍数压缩的下行情景风险。 | 出现流血轮融资,或老股交易价格持续走低并跌破 $1.0B。 |
对照关键尽调结论评估的核心投资驱动因素与下行脆弱点。
[CV004, CV005, CV006, CV008]8.3 融资历程与估值情景推演
CADDi 的融资历程呈现出跨多轮周期的快速估值扩张。2025 年 3 月,公司以 $470 million 估值敲定 $38 million Series C 延期融资,同时借入 $30 million 银行贷款。仅 18 个月后的 2026 年 9 月,公司达成以 $1.2 billion 估值引入 $114 million Series D 融资的确定协议,推动累计股权融资额升至 $234 million。按累计股权融资额计算,当前估值约为其五点一倍,反映出投资人对其未来现金流产出寄予厚望。按 Dealroom 创投对标指标衡量,该轮融资在资金规模上处于第一梯队,但在量化业绩分位数上仅列第 38 百分位左右,很大程度上折射出管理层拒绝公开营收和客户数量的现状。三年情景建模显示:乐观情景下,北美大客户顺利落地带动估值增至 $2.5B 至 $3.0B;基准情景下,依托日本本土核心客群稳固在 $1.2B 至 $1.5B 之间;悲观情景下,一旦国际扩张受阻,估值倍数承压压缩至 $500M 至 $700M。[CV016, CV009, CV010, CV011, CV012]
| 情景 | 核心运营假设 | 隐含估值与回报逻辑 | 下行 / 上行触发因素 | 概率信号 |
|---|---|---|---|---|
| 乐观情景 | 北美业务快速扩张;CADDi Agent 形成平台级粘性;到 2028 年 ARR 达到 $100M+,且软件毛利率达到 80%+。 | IPO 或战略并购时估值达到 $2.5B-$3.0B(相较 Series D 实现 2.0x-2.5x MOIC),由工业 AI 品类领导地位驱动。 | 美国顶尖汽车与工业主机厂加速规模化采购,NRR > 130%。 | 25% |
| 基准情景 | 依托日本企业客群保持平稳增长;北美市场稳步渗透;高昂的客户成功开销使营业利润率维持在温和水平。 | 估值维持在 $1.2B-$1.5B(1.0x-1.25x,保本或微幅收益),反映出稳固的本土壁垒与冗长的海外销售周期并存。 | 本土制造业客群续约率保持稳定,但海外大客户转化周期拉长抵消了部分增速。 | 50% |
| 悲观情景 | 海外大客户渗透遇阻;传统 CAD/PLM 厂商推出原生图纸智能工具;重交付的 CS 模式压缩 SaaS 利润率。 | 估值下调至 $500M-$700M 进行流血轮融资或折价资产出售(相较 Series D 估值缩水 ~40%-60%)。 | 非日本客户出现流失,或未能将重度服务交付转化为标准化的纯软件模式。 | 25% |
基于现有市场与资本结构指标,针对 24-36 个月投资周期的估值情景推演矩阵。
[CV009, CV010, CV011, CV012]8.4 退出准备、逻辑破损触发条件与尽调清单
CADDi 全球员工人数约 900 人,并在芝加哥和东京设立双运营总部,从组织架构上筹划首次公开募股(IPO)或战略并购等长期退出路径。在 Globis、JPS Growth 等日本主力金融机构支持下,东京证券交易所挂牌上市或赴美双重上市是最具可行性的退出路径。Salesforce 等企业软件巨头、丰田等工业财团或工程软件平台发起战略收购,同样是一条可行通路。即便如此,潜在投资人在入场前须严密盯防各项逻辑破损点:估值低于 $1.2 billion 的任何折价轮(down-round)或平价延期融资、联合创始人 Yushiro Kato 与 Aki Kobashi 等高管离职,或是北美获客陷入停滞。排他确认性尽调必须优先索取未经删节的财务审计报告、客户留存队列明细,以及跨优先股层级的完整股权结构表清算优先权条款。[CV007, CV008, CV010, CV013, CV014]
8.5 图表与附录
免责声明
本报告仅供参考,不构成投资建议。
证据索引
| 编号 | 陈述 | 可信度 | 来源 |
|---|---|---|---|
| CO001 | CADDi was founded in Tokyo in 2017 by Yushiro Kato and Aki Kobashi to unleash the potential of manufacturing. | 高 | SO001, SO002 |
| CO002 | CADDi maintains dual corporate headquarters in Chicago, Illinois and Tokyo, Japan. | 高 | SO003, SO004 |
| CO003 | Co-founder and CEO Yushiro Kato previously worked at McKinsey & Company, while co-founder and CTO Aki Kobashi studied at Stanford and led product engineering at Apple. | 高 | SO002, SO004 |
| CO004 | CADDi's executive leadership team includes Chief Financial Officer Ryota Haga and VP of Engineering Chris Cope. | 中 | SO002 |
| CO005 | In September 2026, CADDi entered into definitive agreements for $114 million in Series D funding at a post-money valuation of $1.2 billion. | 高 | SO003, SO004, SO009 |
| CO006 | The Series D round brought CADDi's cumulative lifetime equity funding raised to $234 million. | 高 | SO002, SO003, SO004 |
| CO007 | Prior to Series D, Atomico led a $38 million Series C extension in March 2025 that valued CADDi at $470 million. | 中 | SO004 |
| CO008 | CADDi's Series D brought in new institutional investors Moore Strategic Ventures, Coreline Ventures, HR Tech Fund, Woven Capital, and Salesforce Ventures alongside returning backers Atomico and Globis Capital Partners. | 高 | SO003, SO004 |
| CO009 | CADDi operated CADDi Manufacturing as a custom parts procurement platform from 2017 until officially discontinuing physical parts supply in July 2024. | 中 | SO001 |
| CO010 | In August 2026, CADDi relaunched its software into a vertically integrated architecture comprising a Data Layer, Semantic Layer, Discovery & Agents, and specialized Workflow Products. | 高 | SO001, SO006 |
| CO011 | CADDi reports customer presence across 22 countries and software adoption by more than 50% of discrete manufacturers in the Nikkei 225 index. | 高 | SO001, SO002 |
| CO012 | CADDi's global headcount reached over 910 employees worldwide as of September 2026. | 高 | SO002, SO004 |
| CO013 | Industry critics note that CADDi withholds revenue figures, making its $1.2 billion valuation dependent on growth narratives, while major PLM incumbents like Siemens, SAP, and Dassault could bundle competing AI capabilities. | 中 | SO008 |
| CO014 | Simplify's workforce analytics report CADDi's headcount growth decelerated from previous years, showing 3% growth over 6 months and a 2% decline over 24 months. | 中 | SO008 |
| CO015 | A distinct entity operating under the domain trycaddi.com announced back-office workflow AI agents in August 2026 under CEO Alejandro Castellano, presenting potential trademark and market brand confusion. | 中 | SO007 |
| CM001 | CADDi positions its software as a specialized AI data platform that structures legacy 2D drawings, 3D CAD models, and supply chain records into an enterprise intelligence layer. | 中 | SM001, SM008 |
| CM002 | The target market focuses on discrete manufacturing data utilization while excluding general-purpose document storage and process manufacturing chemical formulation workflows. | 中 | SM004, SM008 |
| CM003 | Status-quo substitutes in manufacturing plants consist of siloed PDM/PLM systems, ERP files, paper drawing archives, and manual folder searches across engineering departments. | 中 | SM006, SM007 |
| CM004 | Discrete manufacturers repeatedly procure near-identical components at disparate prices due to an inability to surface historical drawings and purchasing records. | 中 | SM005, SM008 |
| CM005 | CADDi transitioned from a manufacturing marketplace network to enterprise SaaS in 2022, scaling to more than 180 Japanese enterprise customers and $70 million in ARR by 2024. | 中 | SM003, SM007 |
| CM006 | Over half of Japan's 100 largest discrete manufacturers deploy CADDi software across production facilities to structure technical drawing archives. | 中 | SM008 |
| CM007 | Individual enterprise discrete equipment makers maintain archives ranging from hundreds of thousands to 750,000 legacy engineering drawings spanning decades of custom machinery builds. | 中 | SM005 |
| CM008 | Typical industrial customer projects involve ingesting and normalizing more than 10,000 technical drawings to uncover component pricing inconsistencies and reuse opportunities. | 中 | SM003 |
| CM009 | Primary enterprise software buyers comprise VPs of Procurement, Heads of Engineering, and Chief Information Officers at discrete manufacturing enterprises. | 中 | SM008 |
| CM010 | Day-to-day software users include mechanical design engineers searching for part reuse and quotation specialists calculating custom RFQ pricing against historical cost models. | 中 | SM005, SM006 |
| CM011 | Adoption is triggered by urgent pressures to reduce RFQ turnaround times, win custom machinery bids, and eliminate duplicate part stocking costs across multi-brand operations. | 中 | SM005, SM006 |
| CM012 | Manufacturing enterprises face acute labor shortages and experienced engineering retirements, driving demand for software that digitizes tacit engineering knowledge from drawing archives. | 中 | SM002, SM008 |
| CM013 | Increasing supply chain complexity and tariff volatility force global manufacturers to optimize supplier negotiations and seek alternative component sourcing. | 中 | SM003 |
| CM014 | Organizational change management represents a primary adoption hurdle, requiring engineers to modify long-standing manual design and procurement habits. | 中 | SM008 |
| CM015 | Lack of robust bidirectional API integrations with internal legacy systems creates friction and forces enterprise customers to rely on fragile integration workarounds. | 中 | SM008 |
| CP001 | CADDi positions its Manufacturing AI Data Platform around CADDi Drawer and CADDi Explorer to structure fragmented engineering drawings and procurement data. | 中 | SP005, SP008 |
| CP002 | Third-party software directory Krowdbase lists general document management and collaboration platforms including SharePoint, Box, ONLYOFFICE Docs, and M-Files as primary alternatives to CADDi Drawer. | 中 | SP005 |
| CP003 | General document management and cloud storage solutions lack native computer vision and geometric similarity search for technical engineering drawings. | 中 | SP005, SP006 |
| CP004 | CADDi's software platform utilizes patented similarity search algorithms to match historical 2D drawings by shape, text, dimensions, and parts. | 高 | SP004, SP006 |
| CP005 | Enterprise customers typically manage over 10,000 technical drawings per project, creating significant information retrieval friction in legacy architectures. | 中 | SP007 |
| CP006 | Traditional Product Lifecycle Management (PLM) and ERP systems maintain operational data in organizational silos, requiring manual coordination between engineering and procurement teams. | 高 | SP004, SP007 |
| CP007 | CADDi completed a $114 million Series D funding round at a $1.2 billion valuation in September 2026, bringing total capital raised to $234 million. | 中 | SP008 |
| CP008 | CADDi previously closed a $38 million Series C extension led by Atomico alongside $30 million in bank debt in March 2025. | 中 | SP007 |
| CP009 | CADDi's enterprise customer roster includes industrial conglomerates such as Hitachi, Kawasaki, Subaru, and Tokyo Electron. | 高 | SP007, SP009 |
| CP010 | Stork.AI evaluates Caddi as an AI automation platform across repetitive back-office workflows with freemium tiering for individuals and custom enterprise pricing for teams. | 中 | SP002 |
| CP011 | Enterprise manufacturers face substantial switching costs and data integration overhead when connecting legacy ERP and PLM records to external AI layers. | 中 | SP004, SP008 |
| CP012 | CADDi announced the next-generation CADDi Explorer discovery engine and autonomous CADDi Agent in August 2026 to expand beyond standalone drawing search. | 中 | SP008 |
| CP013 | CADDi is investing in brand marketing and strategic positioning to differentiate its vertical manufacturing data platform against horizontal enterprise software. | 中 | SP003 |
| CP014 | Public enterprise pricing tiers and multi-year contract maintenance commitments for CADDi Drawer remain undisclosed in public marketing surfaces. | 中 | SP002, SP005 |
| CI001 | CADDi completed a $114 million Series D financing round in September 2026 at a post-money valuation of $1.2 billion. | 高 | SI001, SI003, SI004 |
| CI002 | The Series D round lifted CADDi's cumulative equity funding to $234 million, more than doubling its valuation from $470 million in March 2025. | 高 | SI001, SI004 |
| CI003 | CADDi secured $30 million in long-term bank debt from Mizuho and other lending institutions alongside its March 2025 Series C extension. | 中 | SI006 |
| CI004 | CEO Yushiro Kato declined to disclose absolute revenue or customer counts in September 2026 while claiming that sales are more than doubling year over year. | 中 | SI001 |
| CI005 | CADDi operates a dual-stream commercial model providing SaaS software subscriptions alongside custom parts procurement and supply-chain execution services. | 中 | SI002, SI007 |
| CI006 | CADDi Drawer and its successor applications are marketed as software-as-a-service subscriptions targeting annual recurring revenue contracts. | 中 | SI002, SI006 |
| CI007 | CADDi's service delivery model relies on over 100 customer success personnel and forward-deployed engineers to manage customer change management. | 中 | SI001 |
| CI008 | CADDi's total workforce expanded to approximately 900 employees by September 2026, up from 600 staffers in early 2025. | 高 | SI001, SI004 |
| CI009 | CADDi allocates its Series D capital across proprietary AI development, Manufacturing AI Data Platform expansion, North American GTM operations, and technical hiring. | 高 | SI003, SI004 |
| CI010 | Third-party market critiques emphasize that CADDi withholds verified revenue figures, creating an underwriting dependency on uncorroborated growth narratives. | 中 | SI008 |
| CI011 | CADDi markets measurable customer return on investment benchmarks, including a reported 30 percent annual procurement cost reduction at an industrial machinery client. | 中 | SI005 |
| CI012 | CADDi's North American expansion in Chicago targets executive sales and deployment personnel with base salaries ranging between $100,000 and $170,000. | 中 | SI008 |
| CI013 | CADDi reports customer adoption across 22 countries with penetration exceeding 50 percent among Japan's 100 largest manufacturing corporations. | 高 | SI001, SI004 |
| CI014 | Private valuation trackers benchmark CADDi against enterprise supply-chain and data infrastructure comparables commanding high double-digit revenue multiples. | 中 | SI002 |
| CE001 | CADDi organizes its Manufacturing AI Data Platform into four vertical layers spanning data ingestion, semantic modeling, search and discovery agents, and domain-specific workflow applications. | 高 | SE007, SE008 |
| CE002 | In August 2026, CADDi unveiled its next-generation platform architecture comprising CADDi Explorer as the successor to CADDi Drawer, alongside an autonomous CADDi Agent and six value-chain workflow products. | 高 | SE007, SE008 |
| CE003 | CADDi Drawer unifies engineering drawings, assembly hierarchies, parts specifications, and cross-referenced component part numbers into a single searchable repository. | 中 | SE003 |
| CE004 | The platform's proprietary optical character recognition extracts dimensional text, title blocks, and engineering annotations from legacy 2D paper blueprints, scanned PDFs, and handwritten notes. | 中 | SE001 |
| CE005 | CADDi utilizes patented geometric similarity search algorithms to enable engineers and procurement specialists to retrieve historical part drawings based on 2D shape contours and rough sketches. | 中 | SE001 |
| CE006 | CADDi solutions architects and integration engineers construct production customer pipelines using Python, TypeScript, and Rust alongside SQL data orchestration. | 中 | SE004 |
| CE007 | The company's deployment infrastructure utilizes containerized Docker workloads, automated CI/CD pipelines, and multi-cloud hosting environments across AWS, Google Cloud Platform, and Microsoft Azure. | 中 | SE004 |
| CE008 | CADDi's North American enterprise delivery operations require United States citizenship or permanent resident status to comply with International Traffic in Arms Regulations (ITAR) across industrial defense-adjacent supply chains. | 中 | SE004 |
| CE009 | CADDi integrates disparate enterprise data silos by connecting engineering drawings with operational records housed in computer-aided design (CAD), enterprise resource planning (ERP), and product lifecycle management (PLM) systems. | 中 | SE001, SE003 |
| CE010 | By identifying identical and functionally interchangeable components across historical design archives, CADDi facilitates design reuse, standardization, and Value Analysis/Value Engineering (VAVE) cost optimization. | 中 | SE001 |
| CE011 | CADDi's product origins trace from an internal prototype deployed within its physical parts manufacturing business to address the loss of tacit knowledge caused by senior manufacturing engineers retiring. | 中 | SE006 |
| CE012 | CADDi's enterprise engineering challenges include designing software workflows capable of parsing complex mechanical engineering drawings across non-standardized multi-stakeholder use cases. | 中 | SE002 |
| CE013 | The functional effectiveness of CADDi's data platform in production environments is constrained by customer data hygiene, because poorly labeled or corrupted legacy drawings degrade search and recommendation quality. | 中 | SE003 |
| CE014 | CADDi incorporates technical architecture design evaluations into its senior engineering selection process to evaluate candidate decisions regarding database schemas, trade-offs, and cloud infrastructure. | 中 | SE005 |
| CU001 | CADDi's AI data platform serves industrial enterprise customers across automotive, semiconductor equipment, and machinery manufacturing sectors. | 高 | SU005, SU006 |
| CU002 | CADDi reports customer operations spanning 22 countries, serving major industrial manufacturers including Hitachi, Subaru, and Kawasaki. | 中 | SU009 |
| CU003 | Tier-1 automotive supplier Neaton Auto Products Manufacturing expanded CADDi Drawer deployment from an initial 30 users to 85 active accounts, representing a 183% user increase. | 中 | SU001 |
| CU004 | Neaton Auto Products Manufacturing reported a 40% to 70% reduction in drawing search time and a 50% search speed acceleration for quality assurance root cause investigations. | 中 | SU001 |
| CU005 | Heavy equipment supplier Amerequip deployed CADDi Drawer to replace network drive lookups and unify drawing search across enterprise ERP and bill of materials systems. | 中 | SU002 |
| CU006 | Amerequip utilizes CADDi's keyword and similarity search to prevent duplicate part creation and avoid unnecessary purchase orders across legacy inventory. | 中 | SU002 |
| CU007 | A precision metal and plastic manufacturer uploaded over one million documents into CADDi, resulting in tooling estimators handling approximately 75% of RFQ volume on the platform. | 中 | SU003 |
| CU008 | Quoting search durations for sales engineering and tooling estimation decreased from upwards of 30 minutes to one to two minutes following CADDi implementation. | 中 | SU003 |
| CU009 | CADDi recruits post-sales leadership to oversee net revenue retention targets and customer health across complex manufacturing sales cycles. | 中 | SU004 |
| CU010 | Enterprise process automation offerings under the Caddi brand implement tiered SaaS pricing ranging from a free individual tier up to $6,500 monthly for business plans. | 中 | SU007 |
| CU011 | Caddi's online terms of service mandate that separate enterprise order forms or cloud master service agreements supersede standard clickwrap conditions for organizational accounts. | 中 | SU008 |
| CU012 | Industry skeptics highlight that CADDi withholds audited revenue and customer retention metrics while depending on top-line growth narratives across enterprise manufacturing accounts. | 中 | SU009 |
| CU013 | Incumbent PLM and ERP providers such as Siemens, SAP, and Dassault Systemes threaten CADDi's enterprise renewal durability by bundling native AI drawing search into existing software agreements. | 中 | SU009 |
| CU014 | CADDi provides hybrid operations comprising both CADDi Drawer software and CADDi Manufacturing supply chain procurement services. | 高 | SU005, SU006 |
| CR001 | CADDi Drawer introduced ITAR-compliant capabilities to support aerospace and defense manufacturers managing export-controlled design data under U.S. Department of State DDTC regulations. | 中 | SR001, SR003 |
| CR002 | Unlike pharmaceutical compliance platforms such as Veeva, CADDi's drawing management lacks legal mandates requiring platform adoption, leaving its switching moat reliant on cumulative workflow data rather than external regulation. | 中 | SR002 |
| CR003 | Industry analysts and market critics identify a material risk that major enterprise CAD and PLM incumbents like Siemens, SAP, and Dassault Systèmes could bundle equivalent AI-powered drawing search into existing enterprise contracts by 2027. | 中 | SR007, SR002 |
| CR004 | CADDi operates on closed corporate datasets without cross-enterprise drawing sharing, resulting in weak data network effects between distinct customer enterprises. | 中 | SR002 |
| CR005 | CADDi withholds standalone Annual Recurring Revenue (ARR), churn rate, and net revenue retention (NRR) metrics for CADDi Drawer, preventing independent external validation of its underlying SaaS unit economics. | 中 | SR002, SR007 |
| CR006 | In September 2026, CADDi entered into definitive agreements for $114 million in Series D funding at a $1.2 billion valuation, bringing cumulative equity funding to $234 million. | 中 | SR005, SR003 |
| CR007 | In March 2025, alongside its $38 million Series C extension led by Atomico, CADDi secured a $30 million long-term debt facility from Mizuho and other syndication banks. | 中 | SR006 |
| CR008 | Enterprise manufacturing clients frequently rely on legacy software architectures 30 to 40 years old, creating operational friction and data structuring hurdles during initial platform deployment. | 中 | SR002, SR006 |
| CR009 | Despite establishing dual headquarters in Chicago and targeting North American expansion, approximately 77% of CADDi's profiled workforce remains concentrated in Japan. | 中 | SR009 |
| CR010 | CADDi's U.S. go-to-market scaling requires specialized leadership with bilingual English-Japanese fluency to bridge corporate headquarters in Tokyo with North American enterprise operations. | 中 | SR007 |
| CR011 | In July 2024, CADDi terminated its legacy physical parts manufacturing and procurement platform to consolidate resources entirely into its pure software data platform. | 中 | SR002, SR009 |
| CR012 | Because CADDi's platform processes complex geometric tolerances and dimensions for heavy equipment and automotive OEMs, AI extraction errors could lead to defective part sourcing or assembly disruptions. | 中 | SR002, SR004 |
| CR013 | CADDi positions its data platform as a complementary shared intelligence layer rather than a rip-and-replace substitute for existing enterprise PLM and ERP systems. | 中 | SR003 |
| CR014 | CADDi's $1.2 billion valuation is heavily predicated on capturing enterprise accounts in the United States, where enterprise procurement cycles are long and competition from internal data engineering teams is intense. | 中 | SR005, SR007 |
| CR015 | Operating across over 20 countries with major hubs in the U.S., Japan, Vietnam, and Thailand exposes CADDi to divergent regional data residency, labor, and cross-border trade regulations. | 中 | SR003, SR006 |
| CV016 | CADDi's September 2026 financing reinforced a premium valuation and extended the company's capital runway for North American expansion. | 中 | SV004, SV001, SV005 |
| CV002 | CADDi's $1.2 billion valuation in September 2026 more than doubled its previously reported $470 million valuation from March 2025. | 中 | SV001, SV002 |
| CV003 | The Series D financing brought CADDi's cumulative capital raised to $234 million across all equity rounds. | 高 | SV004, SV003 |
| CV004 | CADDi's investor syndicate includes global growth and strategic funds Moore Strategic Ventures, Coreline Ventures, Woven Capital, Salesforce Ventures, Atomico, and Globis Capital Partners. | 高 | SV004, SV003 |
| CV005 | CADDi developed proprietary AI models to parse 2D technical drawings and 3D CAD files while utilizing general-purpose LLMs for textual documents and spreadsheets. | 中 | SV001, SV002, SV006 |
| CV006 | CADDi employs more than 100 customer success staff that outnumber its sales personnel to support hands-on implementation across industrial clients. | 中 | SV001 |
| CV007 | Dealroom benchmarks CADDi's $114 million Series D in the top tier by deal size but ranks it around the 38th percentile against its quantitative score-based benchmark. | 中 | SV002 |
| CV008 | CADDi executive leadership declined to disclose revenue or customer figures while stating that sales were more than doubling year over year. | 中 | SV001, SV002 |
| CV009 | CADDi's global headcount expanded to approximately 900 employees by September 2026, up from 600 in early 2025. | 中 | SV001, SV002 |
| CV010 | More than half of Japan's 100 largest manufacturing companies use CADDi's software platform across their operations. | 中 | SV001, SV002 |
| CV011 | In March 2025, CADDi secured a $38 million Series C extension led by Atomico alongside $30 million in long-term debt financing from Mizuho and other banks. | 中 | SV008 |
| CV012 | CADDi expanded its product suite from CADDi Drawer into CADDi Explorer, CADDi Agent, and six domain-specific manufacturing workflow products. | 中 | SV004, SV001, SV006 |
| CV013 | CADDi maintains operational presence across 22 countries with primary headquarters in Chicago, Illinois and Tokyo, Japan. | 中 | SV001, SV004 |
| CV014 | FundedIQ recorded CADDi's $114 million Series D round as significantly exceeding the $3.9 million median round size for tracked venture rounds in Japan. | 中 | SV009 |
| CV015 | CADDi's $1.2 billion valuation represents a roughly 5.1x ratio against cumulative equity capital raised of $234 million. | 中 | SV001, SV004 |
| 编号 | 出版方 | 标题 | 引文 |
|---|---|---|---|
| SO001 | CADDi | Our History|CADDi | CADDi was founded in 2017 as a procurement platform for custom-manufactured parts... In July 2024, we integrated the two businesses and set a new course as CADDi, the AI Data Platform for Manufacturing... CADDi Manufacturing was discontinued in 2024. We no longer supply parts or assemblies. |
| SO002 | CADDi | Company|CADDi | Yushiro Kato... Co-founder & CEO... Aki Kobashi... Co-founder & CTO... In 2017, he co-founded CADDi with Yushiro Kato. |
| SO003 | Business Wire | CADDi Raises $114 Million in Series D at $1.2 Billion Valuation | CADDi, the global technology company that develops the CADDi Manufacturing AI Data Platform, today announced that it has entered into definitive agreements for $114 million at a $1.2 billion valuation in Series D funding, with eight new and existing investors participating, including Moore Strategic Ventures and Coreline Ventures. This funding brings CADDi's total funding to $234 million. |
| SO004 | SiliconANGLE | CADDi raises $114M at $1.2B valuation to bring manufacturing AI to North America | Since that deal, CADDi's workforce has grown by about half to roughly 900 people while its sales have been more than doubling each year... The company, which now splits its headquarters between Tokyo and Chicago, said it has raised $234 million in all with the Series D included. |
| SO005 | FundedIQ | CADDi: Funding, Investors & Team (Sep 2026) | FundedIQ | CADDi has raised $114M across 1 funding round on record, the most recent a Series D of $114M announced in Sep 2026. |
| SO006 | CADDi | AI Data Platform for Manufacturing|CADDi | A vertically integrated platform that structures and connects fragmented manufacturing data to accelerate execution across every operation. |
| SO007 | PR Newswire | Caddi Launches an AI Agent That Builds and Governs a Firm's Back-Office Agents | The agent discovers a firm's repetitive work, is trained on each process like a new hire, and governs every agent it builds under full audit. |
| SO008 | Simplify | Growth Leadership @ Caddi | Simplify | What critics are saying... Siemens, SAP, and Dassault can bundle similar AI into existing PLM contracts by 2027... CADDi still withholds revenue, making $234 million funding depend on growth narratives, not proof... If U.S. expansion stalls, internal data teams will replicate CADDi using LLMs and warehouses. |
| SO009 | AOL | CADDi Raises $114 Million in Series D at $1.2 Billion Valuation - AOL | CADDi entered into definitive agreements for $114 million at a $1.2 billion valuation in Series D funding. |
| SM001 | CADDi | AI Data Platform for Manufacturing|CADDi | A vertically integrated platform that structures and connects fragmented manufacturing data to accelerate execution across every operation. |
| SM002 | CADDi | AI in Manufacturing turns today's challenges into tomorrow's opportunities|CADDi | CADDi is a digital solutions provider for manufacturers on a mission to "unleash the potential of manufacturing." The company strives to transform the manufacturing industry through its primary offering "CADDi Drawer," an AI-driven data utilization system to assetize data and streamline processes through the patented similarity search function. |
| SM003 | AOL | CADDi Raises $114 Million in Series D at $1.2 Billion Valuation - AOL | The new product, CADDi Explorer, is a manufacturing discovery engine and the next-generation version of CADDi Drawer. |
| SM004 | CADDi | Procurement 101: Demand Planning – Maximize impacts through integration with the supply side|CADDi | If you are in charge of procurement for a manufacturer of industrial machines or plant equipment, that is, products that are high-mix low-volume (HMLV), you might think that the above models are difficult to directly apply because they are designed for mass production. That is indeed the case. |
| SM005 | Business Wire via Yahoo Finance | nVenia Adopts the CADDi AI Data Platform to Assetize 750,000 Drawings and Strengthen Aftermarket Value Across Multiple Brands | nVenia builds highly configured machinery, with 30,000 to 45,000 configuration options on CTO units and fully custom ETO builds. Over decades, that work produced hundreds of thousands of design drawings and a rich, deep component catalog. |
| SM006 | Business Wire via Morningstar | Hyperion Materials & Technologies Partners With CADDi To Reduce Custom Component Quotation Turnaround And Boost Decision Velocity | Hyperion’s engineers handle a high volume of customer requests for quotation (RFQs) that require reviewing technical drawings, evaluating manufacturing feasibility and referencing historical pricing data. |
| SM007 | 3DPrint.com | CADDi Introduces Intelligence to CAD Data Management - 3DPrint.com | Additive Manufacturing Business | Consequently, CADDi pivoted to a SaaS-based software organization, growing its SaaS operations to over 180 customers in Japan, generating $70 million in ARR. |
| SM008 | Rajesh Beri | CADDi — Overview, Features & Use Cases - Rajesh Beri | The VP of Procurement or Head of Engineering at a discrete manufacturer whose drawings, specs, quotes and defect history live in separate systems, so the same part gets redesigned and re-sourced instead of reused. |
| SP002 | Stork.AI | Caddi Review (2026): Pricing & Alternatives | Caddi is an AI-powered platform developed by Caddi (company) that enables manufacturing companies, wealth management firms, law firms, and insurance companies to automate repetitive back-office tasks. |
| SP003 | Built In Chicago | Brand Marketing Manager - CADDi | Develop and champion long-term brand positioning across core markets and channels, ensuring alignment with overall business goals. Conduct comprehensive brand audits, competitive analysis, and refine CADDi's brand architecture. |
| SP004 | CADDi | VOICES: Philip Carpenito, Strategic Advisor at CADDi - Design World | Better visibility into historical engineering, procurement and operational data enables faster, lower-risk, more cost-effective sourcing decisions. Rather than treating each product as a new effort, CADDi software can help companies leverage institutional knowledge, avoid repeating mistakes and capitalize on proven solutions. |
| SP005 | Krowdbase | Top 13 CADDi Drawer Alternatives & Competitors - 2026 - Krowdbase | Some of the notable alternatives to CADDi Drawer include ONLYOFFICE Docs, Microsoft SharePoint, M-Files, DocuWare, Laserfiche, OpenText Content Management, OnBase, Box, Dropbox Business, Revver, Zoho WorkDrive, Adobe Acrobat, Google Drive. |
| SP006 | CADDi | AI in Manufacturing turns today's challenges into tomorrow's opportunities|CADDi | CADDi Drawer is designed to highlight operational insights, which is especially valuable for manufacturers dealing with multiple product lines or high-output facilities. |
| SP007 | CADDi | CADDi raises $38M Series C Extension led by Atomico to double down on rapid growth of their AI data platform for manufacturing companies | One example is to optimise parts pricing through a similarity search of historical technical drawings to find pricing inconsistencies. For an average customer, one project typically involves over 10,000 individual technical drawings. |
| SP008 | AOL | CADDi Raises $114 Million in Series D at $1.2 Billion Valuation - AOL | The new product, CADDi Explorer, is a manufacturing discovery engine and the next-generation version of CADDi Drawer. CADDi Agent is a new AI agent that analyzes manufacturing data, makes decisions, and takes action within each company's operating context. |
| SP009 | Simplify | Growth Leadership @ Caddi | Simplify | CADDi builds a Manufacturing AI Data Platform that organizes fragmented data from engineering, procurement, production, and quality so humans and AI can reason across decades of drawings and supply-chain records. |
| SI001 | Fortune | Exclusive: Manufacturing AI startup CADDi valued at $1.2 billion in $114 million funding round | Fortune | Kato declined to disclose revenue or customer numbers, but said sales are more than doubling year over year and that the company now has customers in 22 countries, although the U.S. is a core focus. |
| SI002 | Multiples | CADDi Valuation, Funding & Investors | Multiples | CADDi reports remarkable revenue growth exceeding T2D3 SaaS benchmarks and immediate ROI for clients, such as one automotive parts customer reducing fastener SKUs by 60%, targeting $1 billion in software revenue by 2030. |
| SI003 | CADDi | CADDi Raises $114 Million To Fix Manufacturing’s Biggest Bottleneck With AI|CADDi | CADDi, dual-headquartered in Tokyo and Chicago, just raised $114 million in Series D funding. The round values the company at $1.2 billion. That’s up from $470 million in March 2025. Total funding is now $234 million. |
| SI004 | Business Wire via Yahoo Finance | CADDi Raises $114 Million in Series D at $1.2 Billion Valuation | CADDi, the global technology company that develops the CADDi Manufacturing AI Data Platform, today announced that it has entered into definitive agreements for $114 million at a $1.2 billion valuation in Series D funding, with eight new and existing investors participating, including Moore Strategic Ventures and Coreline Ventures. This funding brings CADDi's total funding to $234 million. |
| SI005 | CADDi | AI Data Platform for Manufacturing|CADDi | A vertically integrated platform that structures and connects fragmented manufacturing data to accelerate execution across every operation. |
| SI006 | CADDi | CADDi raises $38M Series C Extension led by Atomico to double down on rapid growth of their AI data platform for manufacturing companies|CADDi | CADDi, the AI data platform for manufacturing companies on a mission to unleash the potential of manufacturing through democratising supply chain data, today announced the closing of a $38M Series C Extension financing, led by Atomico. The company additionally secured $30M long-term debt from Mizuho and other banks. |
| SI007 | EquityZen | Invest In CADDi Stock | Buy Pre-IPO Shares | EquityZen | CADDi is an online ordering platform that tracks industrial products for the manufacturing business. They provide Realize DX throughout the supply chain by offering both goods and software. It enables businesses to reduce procurement expenses, improve QCD, and expand capacity. |
| SI008 | Simplify | Growth Leadership @ Caddi | Simplify | What critics are saying ... Siemens, SAP, and Dassault can bundle similar AI into existing PLM contracts by 2027. ... CADDi still withholds revenue, making $234 million funding depend on growth narratives, not proof. ... If U.S. expansion stalls, internal data teams will replicate CADDi using LLMs and warehouses. |
| SE001 | CADDi | Design for Manufacturability: Building Better Products from the Ground Up|CADDi | Its best-in-class OCR-based AI technology extracts data from legacy 2D paper drawings, scanned PDFs, and even handwritten notes. The core innovation lies in its patented similarity search technology, which allows users to instantly find and compare past designs based on their shape and features, even from a hand-drawn sketch. |
| SE002 | CADDi Tech Careers | Product Manager | The challenge of complex product design built around engineering drawings — the core data of manufacturing — across diverse stakeholders and use cases |
| SE003 | CADDi | Amerequip's CADDi Seizes Industrial AI's Data Bottleneck as S-Curve Steepens|CADDi | CADDi Drawer is built as a first-principles solution to this. It creates a single searchable environment that unifies engineering and quality teams by connecting drawings, assemblies, specifications, and cross-referenced part numbers into one accessible system. |
| SE004 | Built In Chicago | Senior Solutions Architect Expert- (Data Science) - CADDi | Code and ship integrations in Python/TypeScript/Rust, build APIs and integrations, and stand up data pipelines. |
| SE005 | Scribd | Backend Architecture Work Sample_EN | This is a technical architecture work sample intended for experienced engineers, where we ask you to write a design document for a hypothetical situation. |
| SE006 | CADDi Tech Blog | Building for the future in 2025 - CADDi Tech Blog | What started out for us as just a simple demo, implemented as a grid of thumbnails powered by an off the shelf search tool, has now morphed into a powerful knowledge engine for some of the largest corporations in Japan and abroad. |
| SE007 | CADDi | AI Data Platform for Manufacturing|CADDi | A vertically integrated platform that structures and connects fragmented manufacturing data to accelerate execution across every operation. |
| SE008 | AOL | CADDi Raises $114 Million in Series D at $1.2 Billion Valuation - AOL | The new product, CADDi Explorer, is a manufacturing discovery engine and the next-generation version of CADDi Drawer. CADDi Agent is a new AI agent that analyzes manufacturing data, makes decisions, and takes action within each company's operating context. CADDi will also offer six domain-specific Workflow Products spanning the manufacturing value chain, enhancing decision-making and execution to drive business-critical impact. |
| SU001 | CADDi | Neaton Auto Products Manufacturing (NAPM) improved quality control and reduced search time by 70% with CADDi AI data platform | CADDi platform user adoption increased from an initial 30 users to 85 active accounts as usage of CADDi spread, reflecting a 183% increase as usage expanded across teams. |
| SU002 | CADDi | Amerequip CADDi Case Study | We don’t need to go to LX, do a ‘where used’ BOM, then open the drawing server and search the new part anymore. It’s very handy. |
| SU003 | Business Wire | CADDi Enables a Leading Manufacturer of High-Precision Metal and Plastic Products to Decrease Quoting Time by 96% | The manufacturer’s tooling estimators now rely on CADDi for approximately 75 percent of their RFQ volume. |
| SU004 | Built In | Director of Customer Success - CADDi | A proven track record of scaling post-sales or services organizations, optimizing NRR targets, and defining departmental metrics. |
| SU005 | CADDi | AI Data Platform for Manufacturing | Decisions and outcomes are automatically captured as data, creating a continuous cycle of knowledge creation and use. |
| SU006 | CADDi | AI in Manufacturing turns today's challenges into tomorrow's opportunities | CADDi also supports manufacturers through "CADDi Manufacturing," the procurement and manufacturing service that optimizes quality, cost, and delivery within supply chain infrastructure. |
| SU007 | Caddi Technologies | Caddi vs. Celonis: Find the Work, Then Build It (2026) | Individual $0. Team $250 per user per month. Business $6,500 per month. Enterprise custom. |
| SU008 | Caddi Technologies | Terms of Use | Caddi | If you have entered into a separate written agreement with us governing your use of the Services, including a Cloud Service Agreement, Master Services Agreement, Enterprise Order Form, or Data Processing Addendum, that agreement controls to the extent of any conflict with these Legal Terms. |
| SU009 | Simplify | Growth Leadership @ Caddi | Simplify | What critics are saying ... Siemens, SAP, and Dassault can bundle similar AI into existing PLM contracts by 2027. ... CADDi still withholds revenue, making $234 million funding depend on growth narratives, not proof. ... If U.S. expansion stalls, internal data teams will replicate CADDi using LLMs and warehouses. |
| SR001 | Govly | CADDi Launches ITAR Support for Aerospace Manufacturers | Govly | CADDi Co. Ltd. has introduced ITAR-compliant capabilities within its CADDi Drawer software platform to support aerospace and defense manufacturers managing sensitive design data regulated by the U.S. Department of State's Directorate of Defense Trade Controls (DDTC). This enhancement enables secure handling of classified information with strict U.S. data residency and access controls, facilitating compliance with export control regulations while supporting AI-driven innovation in aerospace manufacturing. |
| SR002 | Stockton / GTM Strategy | CADDi × Veeva Systems - Can the 'Mechanism That Created a 5.1 Trillion Yen Pharmaceutical Company' Be Replicated in Manufacturing?|ストックトン / GTM戦略 | However, the ARR (Annual Recurring Revenue), churn rate, and NRR (Net Revenue Retention) of Drawer alone are not public, so the quality of its revenue as a SaaS cannot be verified from the outside. |
| SR003 | Automate Show | AI Data Platform for Manufacturers. | CADDi Co. Ltd. CADDi Inc. is an international technology company offering an AI-powered data intelligence platform purposefully built for design, engineering and manufacturing businesses. We unleash manufacturing potential by transforming years of information into actionable data intelligence across the discrete manufacturing supply chain. |
| SR004 | CADDi | AI in Manufacturing turns today's challenges into tomorrow's opportunities|CADDi | The company strives to transform the manufacturing industry through its primary offering "CADDi Drawer," an AI-driven data utilization system to assetize data and streamline processes through the patented similarity search function. |
| SR005 | AOL / Business Wire | CADDi Raises $114 Million in Series D at $1.2 Billion Valuation - AOL | CADDi, the global technology company that develops the CADDi Manufacturing AI Data Platform, today announced that it has entered into definitive agreements for $114 million at a $1.2 billion valuation in Series D funding, with eight new and existing investors participating, including Moore Strategic Ventures and Coreline Ventures. This funding brings CADDi's total funding to $234 million. |
| SR006 | CADDi | CADDi raises $38M Series C Extension led by Atomico to double down on rapid growth of their AI data platform for manufacturing companies|CADDi | CADDi, the AI data platform for manufacturing companies on a mission to unleash the potential of manufacturing through democratising supply chain data, today announced the closing of a $38M Series C Extension financing, led by Atomico. The company additionally secured $30M long-term debt from Mizuho and other banks. |
| SR007 | Simplify | Growth Leadership @ Caddi | Simplify | Siemens, SAP, and Dassault can bundle similar AI into existing PLM contracts by 2027. CADDi still withholds revenue, making $234 million funding depend on growth narratives, not proof. If U.S. expansion stalls, internal data teams will replicate CADDi using LLMs and warehouses. |
| SR009 | Dealroom | CADDi — Colt company profile | Dealroom | The company initially operated as a tech-driven manufacturing marketplace for custom parts before shifting its focus to a SaaS model three years ago. Its core offerings are now 'CADDi Drawer', an AI-powered drawing management system, and 'CADDi Quote', a transformative quoting solution. |
| SV001 | MasterNode AI | CADDi raises $114M Series D at $1.2B valuation for manufacturing AI | The valuation is more than double the $470 million the company reported in March 2025. |
| SV002 | Dealroom | CADDi hits $1.2B valuation with $110 million Series D | Dealroom News | The round sits in the top tier of recent deals by size, ranking around the 38th percentile against a score-based benchmark. |
| SV003 | Pulse 2.0 | CADDi Raises $114 Million Series D At $1.2 Billion Valuation | The financing increases CADDi's total funding to $234 million. New investors in the round include Moore Strategic Ventures, Coreline Ventures, HR Tech Fund, Woven Capital and Salesforce Ventures. |
| SV004 | CADDi | CADDi Raises $114 Million in Series D at $1.2 Billion Valuation|CADDi | CADDi, the global technology company that develops the CADDi Manufacturing AI Data Platform, today announced that it has entered into definitive agreements for $114 million at a $1.2 billion valuation in Series D funding, with eight new and existing investors participating, including Moore Strategic Ventures and Coreline Ventures. |
| SV005 | Business Wire via Yahoo Finance | CADDi Raises $114 Million in Series D at $1.2 Billion Valuation | This is a paid press release. Contact the press release distributor directly with any inquiries. |
| SV006 | AOL | CADDi Raises $114 Million in Series D at $1.2 Billion Valuation - AOL | The new product, CADDi Explorer, is a manufacturing discovery engine and the next-generation version of CADDi Drawer. |
| SV008 | CADDi | CADDi raises $38M Series C Extension led by Atomico to double down on rapid growth of their AI data platform for manufacturing companies|CADDi | CADDi, the AI data platform for manufacturing companies on a mission to unleash the potential of manufacturing through democratising supply chain data, today announced the closing of a $38M Series C Extension financing, led by Atomico. The company additionally secured $30M long-term debt from Mizuho and other banks. |
| SV009 | FundedIQ | CADDi: Funding, Investors & Team (Sep 2026) | FundedIQ | CADDi has raised $114M across 1 funding round on record, the most recent a Series D of $114M announced in Sep 2026. That round is above the $3.9M median round size for Japan companies tracked on FundedIQ over the last two years. |