UltraGreen.ai

1.1 Identity, stage, and platform

UltraGreen.ai is best grounded as a Singapore-headquartered, publicly listed healthcare company focused on fluorescence-guided surgery rather than as a generic AI-health story. Independent 2026 analyst coverage and September 2025 reporting both describe the company as Singapore-based and centered on fluorescence-guided surgical technologies, while the official product pages frame the commercial stack around ICG dyes, IC-Flow imaging hardware, PerfusionWorks quantification software, and UltraGreen Data Systems. That matters for the rest of the report because the identity surface is broad, but the retained evidence still points to a concrete operating core: dye-led workflows supported by hardware, analytics, and cloud tooling. The retained official sources are useful for product architecture and positioning, but they should be treated as company claims unless matched by independent material.[CO001, CO002, CO003, CO004]

1.2 Leadership, governance, and key-person dependence

The named leadership bench is strong on paper for a newly listed medtech issuer. Ravi Sajwan remains central as founder and CEO, and the analyst material ties him to multiple prior startup and CTO roles, which supports founder-market-fit on technical and company-building dimensions. The board chair role is held by Kwa Chong Seng, whose background spans ExxonMobil Asia Pacific, DBS, Temasek, and Singapore Exchange, giving the company established public-market oversight credentials. Even so, the sharpest independent concern is not résumé weakness but post-IPO discipline: ICAM explicitly argues that the real governance question is whether management reinforces the proven ICG franchise or broadens into software, data, and adjacent initiatives faster than the economics justify. That makes Sajwan concentration, capital allocation, and strategic scope the main governance watchpoints rather than distress or obvious control abuse.[CO005, CO006, CO014, CO015]

1.3 Capitalization, scale signals, and milestone chronology

The strongest recent balance-sheet signal is the September 2025 strategic investment, which multiple retained sources tie to a US$188 million raise at a US$1.3 billion equity valuation led by 65 Equity Partners and Vitruvian, with August Global Partners participating. Independent January 2026 coverage also shows a concentrated shareholder base led by Renew Group, plus Anchor VI and Verde Taano, which is relevant for both governance and future float dynamics. On operating scale, the cleanest retained revenue datapoint is the January 2026 analyst model that lists 2025A revenue of US$142 million versus US$115 million in 2024A. The chapter also surfaces an important disclosure inconsistency: official and analyst materials do not agree on country approvals and commercial reach, citing 35+ approved markets, 54 countries sold, or 40 approvals and 55+ countries sold. That inconsistency is material because later market, customer, and valuation arguments depend on the precise footprint definition.[CO007, CO008, CO009, CO010, CO011, CO012, CO013]

1.4 Exhibits

2.1 Market boundary: fluorescence-guided surgery, not all surgery

UltraGreen’s relevant market is the fluorescence-guided surgery workflow, not general hospital IT or all surgical spending. The company’s own positioning combines an ICG fluorescence agent with near-infrared imaging, quantification software, and AI-enabled analytics, while independent coverage still shows that the recurring economic core is the consumable dye used during procedures. That matters for valuation: most surgical spend remains outside the addressable pool unless a procedure uses fluorescence, the site has compatible imaging, and the workflow supports repeat ICG usage. The closest status-quo substitute is still white-light or non-fluorescence surgery, while the closest adjacent alternatives are hardware-bundled fluorescence offerings that may not compete for every hospital account. This boundary keeps the chapter from treating “surgical AI” as synonymous with UltraGreen’s addressable market. [CM001, CM002, CM003, CM010]

2.2 Sizing lenses point to runway, but not a clean public SAM/SOM

The strongest public evidence supports multiple sizing lenses rather than a single clean TAM/SAM/SOM stack. Independent analysis citing Frost & Sullivan says U.S. fluorescence-guided surgery penetration remains below 25% of eligible procedures, while Asia-Pacific is forecast to be the fastest-growing regional ICG market through 2030. Company and IPO reporting add a second lens: UltraGreen had sold dye products in 54 countries from 2022 through the first half of 2025 and shipped 5.3 million vials since 2015, which shows meaningful real-world adoption but not the total remaining opportunity. A third lens comes from management commentary relayed by ValueZoomer, which frames 10 million annual U.S. addressable procedures against 670,000 vials sold last year. That signals headroom, but it is still not a reconciled, auditable SAM/SOM model by geography, price, attach rate, and product bundle. [CM004, CM005, CM006, CM007, CM008, CM011, CM015]

2.3 Buyers, users, and adoption path

The end user is the surgeon and operating team, but adoption and monetization move through a more layered buyer path. UltraGreen’s sources emphasize surgeon education, KOL relationships, dosing guidance, and workflow standardization because clinical pull is required before a hospital expands fluorescence usage. Commercial conversion, however, still depends on hospital procurement, reimbursement logic, installed imaging systems, and distributor execution in local markets. The buyer map also differs by product layer: dye purchasing behaves like a recurring procedural consumable, imaging behaves more like capital equipment access, and software depends on whether hospitals want quantitative reporting and data integration. This explains why the market can be clinically broad but commercially uneven across countries, specialties, and account types. [CM009, CM010, CM012, CM013]

2.4 Constraints and unresolved diligence gaps

The key market constraint is not whether ICG works in principle; it is how quickly hospitals, regulators, and competing suppliers convert that clinical utility into repeatable commercial adoption. DBS’s March 2026 note makes that visible in APAC: imaging-system approvals expanded, but the dye remained registered only in the Philippines among the newly approved territories, keeping near-term revenue contribution small. Competitive entry risk is also real. UltraGreen’s August 2026 shareholder update confirms new U.S. approvals for rival suppliers, even as management argues that approval alone does not prove launch or penetration. Finally, public footprint statistics are not yet normalized across “registered,” “approved,” “sold,” “exemption,” and “regulatory footprint” definitions, so any market-share or adoption model built from those numbers still needs management reconciliation before it can support a valuation-grade market map. [CM008, CM014, CM015]

3.1 Competitive set is a workflow stack, not one direct SKU peer

The retained evidence points to a layered competitive landscape rather than a clean one-company peer set. UltraGreen describes itself as combining a trusted ICG agent, fluorescence imaging, quantification software, and AI-enabled analytics, while independent commentary frames the same stack as three value layers: consumables, imaging systems, and software. That means the company is exposed to several alternative solution classes at once: regional ICG suppliers, installed imaging platforms that can embed fluorescence capability, stand-alone perfusion quantification tools, and status-quo non-fluorescence workflows. The competitive question is therefore whether hospitals buy into a broader data-backed fluorescence workflow or stay with narrower procedural substitutes that solve only one step.[CP001, CP005, CP012, CP013, CP014]

3.2 The moat is strongest in consumables scale and regulatory breadth

The strongest evidence for defensibility sits in ICG consumables rather than hardware alone. FSM Global says UltraGreen held approximately 68% of global ICG vial volume and 63% of industry revenue in 2024, with especially strong US and European share, and it describes competing manufacturers as more regional with narrower approvals and distribution. The same report argues that approvals can take years, and that hospital procurement relationships and surgeon familiarity make displacement slow once a vendor is embedded. UltraGreen's own 2026 update reinforces the breadth point by citing 41 Verdye countries and 45 IC-Flow markets, which widens the installed territory a smaller rival would need to match.[CP002, CP003, CP004, CP005, CP011]

3.3 Workflow breadth helps, but substitutes still win on some procedures

The acquisition of Perfusion Tech matters because it shifts the comparison from image visibility alone toward objective perfusion quantification and structured workflow data. Biospectrum Asia and Perfusion Tech both describe the combined offering as moving toward standardised, reproducible decision support, while UltraGreen's own pipeline language explicitly frames the goal as turning qualitative imaging into quantified insight. Even so, substitute workflows remain credible. In early-breast-cancer sentinel-node mapping, the Frontiers study reported a 90.6% per-patient detection rate for ICG alone versus 97.1% for Tc-99m and 98.1% for the combined technique, showing that legacy nuclear-medicine pathways can still outperform on specific tasks even when ICG has workflow and safety advantages.[CP006, CP007, CP009, CP010, CP012]

3.4 Durability rests on adoption breadth, but evidence is not universal

The chapter's best positive case is that UltraGreen can pair recurring per-procedure consumables with expanding imaging and software layers, while robotics adoption and new approvals increase the number of sites able to use fluorescence-guided workflows. The main counterpoint is that the clinical and operational case is uneven across procedures. The SAGES systematic review says FIGS with ICG improves some applications, especially colorectal leak reduction, but that evidence remains limited or context-specific elsewhere and access is still constrained in many settings. That leaves the moat real but not absolute: the company appears differentiated where buyers value a full fluorescence stack, yet procurement, pricing, attach-rate, and procedure-by-procedure superiority data remain missing from public evidence.[CP008, CP010, CP011, CP013, CP014]

3.5 Exhibits

4.1 Consumables drive the disclosed economics

Public evidence supports a consumables-led revenue model rather than a balanced three-line mix. UltraGreen’s investor page reported FY2025 revenue of US$142.4 million and adjusted EBITDA of US$89.4 million, while the same investor surface and third-party summaries put FY2025 gross margin at roughly 85%. Independent analyst coverage described the current business as mainly driven by recurring ICG consumables, which matters because recurring dye usage is the only clearly evidenced engine converting procedure volume into repeat revenue today. Official company materials also position imaging hardware, quantification software, and AI analytics as part of the platform, but the assigned evidence does not disclose separate revenue contribution, ASPs, attach rates, or paying-software penetration. That leaves hardware and software as strategically important but financially under-specified layers on top of a strong consumables core.[CI001, CI002, CI003, CI004, CI005]

4.2 Capital exists, but forward earnings conversion is still conditional

The public record shows multiple capital milestones, but not a fully underwritable runway model. At IPO launch, press coverage said UltraGreen marketed about 112 million shares at US$1.45, with roughly US$150 million of gross proceeds due to the company; management also framed the proceeds as funding core product development, geographic expansion, and working capital. Before listing, Allen & Gledhill described an October 2025 secondary sale at a US$1.3 billion equity valuation, and a later analyst summary reported FY2025 net cash of US$173.0 million. That means capital adequacy is better supported than burn or runway, yet the next leg of earnings is not frictionless: Tickrs highlighted a sharp rise in effective tax rate, and DBS argued new APAC device approvals mainly create option value until the core ICG dye is approved in the same markets and revenue scales beyond a small base.[CI006, CI007, CI008, CI009, CI010, CI013, CI014, CI015]

4.3 Verdict: high-quality gross margin, low transparency on unit economics

The best-supported conclusion is that UltraGreen already operates a profitable, high-margin model anchored in recurring ICG consumables and global geographic reach. What is not public is equally important: the assigned sources do not provide realized dye pricing, hardware ASPs, software contract structure, CAC, payback, sales-cycle duration, customer concentration, distributor concentration, or a bridge from net cash to runway. The latest disclosed share price, materially below the IPO price, does not prove a funding shortfall, but it does suggest any future equity financing would happen from a weaker market position than at listing. For diligence, the chapter should therefore be read as positive on revenue quality and margin structure, but cautious on forward underwriting because capital intensity and commercial efficiency are still opaque in public materials.[CI003, CI004, CI010, CI016, CI017]

4.4 Exhibits

5.1 Integrated product stack

The product set in the assigned pool is broader than a single indocyanine green vial. Company-controlled and third-party materials consistently describe a stack that starts with ICG consumables, extends into near-infrared imaging hardware, and aims to add quantification software and data infrastructure around the surgical workflow. The official May 2026 Verdye release says the company is building a "comprehensive fluorescence-guided surgery ecosystem" that combines pharmaceutical products, compatible imaging hardware, and AI-powered data platforms, while the prospectus and analyst coverage describe the same architecture as a vertically integrated model rather than a pure consumables business. That matters because it changes both the moat narrative and the diligence surface: the business is not only selling dye into existing procedures, it is trying to influence clinical workflow, imaging capture, interpretation, and eventually structured post-operative data. The stack is also segmented by maturity. The consumable layer is the most established, with IC-GREEN in the US and Verdye outside the Americas positioned as the core clinical inputs. Imaging hardware appears commercially tangible as well: independent coverage and company-linked approval notices both describe IC-Flow V2 as a handheld near-infrared system used intraoperatively for perfusion and lymphatic visualization. The software and data layers look earlier and more strategic than proven at scale. Public materials emphasize PerfusionWorks, quantification, structured reports, and AI-powered insights, but the assigned pool provides much less implementation detail on integration depth, installed-base adoption, release cadence, or support ownership than it does on the dye and device layers. [CE001, CE002, CE003, CE005, CE012]

5.2 Workflow and operating architecture

The operating model implied by the assigned evidence follows the intraoperative workflow. ICG is administered using procedure-specific timing and dosing protocols; near-infrared imaging then captures the fluorescence signal; quantification software is meant to reduce subjective interpretation; and a cloud/data layer is intended to turn fluorescence observations into structured reports and reusable datasets. The ISFGS dosing chart is particularly useful here because it shows that clinical use depends on disciplined protocol choices by procedure, dose, route, and timing rather than on a generic "AI surgery" abstraction. In other words, Ultragreen's software ambitions sit on top of a clinical workflow that already has technique-specific operating constraints. The software ambition is most explicit in the company's product pages and the Perfusion Tech acquisition announcement. The Innovation Pipeline page says the company is "moving qualitative assessment to quantitative, standardized insights" and building a cloud platform that integrates with hospital systems. The clinical-research page adds that PerfusionWorks and UltraGreen Data Systems are intended to turn fluorescence into objective metrics and post-operative review capability, while the Perfusion Tech announcement describes real-time visualization, quantitative analytics, cloud-based data management, and AI-powered insights as one platform. That set of claims is coherent enough to author an architecture figure, but not detailed enough to validate data models, deployment models, or software assurance controls from public evidence alone. [CE003, CE004, CE005, CE006, CE007]

5.3 Maturity, approvals, and technical risk

Product maturity is uneven across layers and geographies. On the positive side, assigned evidence supports real commercial and regulatory progress: Verdye won Singapore approval in May 2026, while a syndicated approval notice and DBS research say IC-Flow V2 approvals in India, Thailand, the Philippines, and Bangladesh pushed the imaging platform to 45 territories. The August 2026 shareholder update also shows management trying to reposition the story beyond a single-dye franchise by highlighting dye, imaging, and software together and by saying PerfusionWorks completed MDR conformity assessment while awaiting formal CE certification. The main risk is that the software-and-platform story is more legible strategically than it is operationally. FSMOne says PerfusionWorks has FDA 510(k) clearance, but management's later August 2026 update says formal CE certification is still pending and that FDA clearance would be pursued afterward; the discrepancy requires primary regulatory documentation. DBS also notes that several APAC imaging approvals do not yet imply corresponding dye approvals, so the full platform cannot be assumed to land as one synchronized bundle. The public evidence is thin on trust and reliability surfaces that matter for hospital software diligence: the assigned pool does not document security certifications, uptime or status reporting, interface standards, or public practitioner/developer telemetry around the data layer. Finally, the geography story needs reconciliation because materials variously refer to more than 43 countries in regulatory footprint, 45 imaging territories, and more than 55 countries of product sales. Product-by-product maturity should therefore be underwritten with reconciled regulatory and commercial coverage rather than headline geography counts. [CE008, CE009, CE010, CE011, CE013, CE014, CE015, CE016]

5.4 Exhibits

6.1 The commercial base is a procedure-linked healthcare ecosystem

UltraGreen.ai serves a multi-party healthcare workflow rather than a conventional software account. Surgeons and operating-room teams use ICG fluorescence during procedures; hospitals and health systems control procurement and budgets; and pharmaceutical wholesalers connect the company to institutional ordering channels. The company also presents imaging hardware and data tools as extensions around the consumable. That segmentation matters because the recurring event is a procedure, not a seat renewal. Current customer quality should therefore be judged through vial use, institutional access, supply reliability, and clinical workflow adoption. Public materials describe the intended beneficiaries and platform logic, but they do not disclose revenue by buyer type, payer, hospital size, or individual account. [CU001, CU002, CU003, CU004]

6.2 Usage growth is visible, but active-account depth is not

Independent coverage provides the clearest adoption signal: FY2025 ICG volume reached 987.7 thousand vials, up from 872.9 thousand in FY2024, while both volume and average selling price increased. Revenue growth also appeared in both US and non-US markets. These measures show that commercial momentum was not solely a pricing effect, although vial growth cannot distinguish more procedures at existing sites from new account wins. Regulatory reach is also expanding, but the source set uses different denominators for dye approvals, imaging-system clearances, overall footprint, and countries sold into. Those measures should not be collapsed into a single customer count, and none substitutes for active hospitals, utilization per site, or repeat-purchase cohorts. [CU005, CU006, CU007, CU010, CU011]

6.3 Named proof stops at channel counterparties

The retained sources name Cencora, McKesson, and Cardinal Health as major US wholesalers, but they do not name production hospitals, health systems, or surgeon reference accounts. The three wholesalers represented 68.2% of revenue in 2024 according to DBS and 67.8% in FY2025 according to The SEA Analyst. That persistence supports broad channel access, yet it is a concentration statistic at the intermediary layer rather than proof of a diversified end-customer base. Individual wholesaler shares, downstream hospital concentration, contract duration, and switching behavior remain undisclosed. The result is strong distribution evidence but weak account-level reference quality: public diligence can verify that product moves through scaled channels, not which institutions use it, at what intensity, or with what renewal behavior. [CU008, CU009, CU012]

6.4 Clinical utility supports repeat use, while conversion and retention remain open

Independent clinical evidence strengthens the rationale for hospital adoption in selected procedures. The SAGES systematic review reported lower colorectal anastomotic leak rates with ICG-guided surgery, while also warning that evidence remains limited in several other applications. Commercial durability still requires a separate proof layer. UltraGreen.ai itself cautions that regulatory approval does not establish launch, customer adoption, or market penetration, and independent analysis notes that surgeons and hospitals can change practice slowly. The credible expansion loop is therefore evidence to approval, institutional procurement, procedure-level repeat use, and only then hardware or data-platform attachment. Public sources support each strategic surface, but they do not quantify conversion rates, NRR, GRR, churn, renewals, contracts, or cross-product attach rates. [CU013, CU014, CU015, CU016]

6.5 Exhibits

7.1 Severity ranking: approvals, manufacturing, and concentration drive the residual risk stack

The current evidence points to a three-layer risk stack. First, regulatory timing remains the most material near-term exposure because UltraGreen can open device pathways ahead of pharmaceutical registrations, leaving commercial readiness uneven by country even after a favorable headline on geographic expansion. Second, the company's core dye business still carries manufacturing and logistics fragility: management described ICG production as "extremely complex," and independent reporting says Ireland remains the manufacturing base while the US is still the largest market. Third, the equity story still depends on extending a consumables-led franchise into a broader software and data platform without public evidence yet proving customer concentration, attach rates, or security-control maturity for the cloud roadmap. [CR002, CR003, CR004, CR007, CR008, CR010, CR011, CR016]

7.2 Regulatory risk is declining, but still sequenced market by market

Regulatory progress is real: management disclosed a May 2026 Singapore Verdye approval and a 41-country / 45-market installed approval footprint, while the company's March 2026 news archive highlighted new IC-Flow approvals across South and Southeast Asia. That lowers the probability of a broad platform rejection. The residual issue is sequencing. DBS explicitly warned that several new device-approval markets still lacked the matching dye approval or exemption pathway, which means footprint expansion does not automatically translate into near-term revenue. The same evidence base also leaves open a legal-visibility gap: the assigned pool did not surface readable prospectus risk factors, litigation detail, or privacy/security filings, so investors still need a direct public-record review before treating the regulatory story as fully de-risked. [CR001, CR002, CR003, CR004, CR015]

7.3 Operational scaling depends on manufacturing reliability and surgeon-led distribution loops

UltraGreen's operating model depends on more than regulatory wins. Company materials say fluorescence adoption remains fragmented by tooling, training, and integrated-platform access, and the innovation-pipeline page says expansion into new procedures comes from close relationships with pioneering surgeons and long-term clinical partnerships. That creates a credible adoption engine, but it also means the company depends on a relatively concentrated external ecosystem of clinicians, hospitals, and channel partners to move from pilot enthusiasm to routine procedural volume. The model is further concentrated economically: independent initiation coverage says the business is still mainly driven by recurring ICG consumables, while the company's AI and data-platform upside remains an execution layer rather than the already-proven revenue base. [CR005, CR006, CR010, CR011, CR012]

7.4 Mitigations exist, but the thesis should break if commercial proof does not catch up

The mitigation case is meaningful but conditional. UltraGreen announced a US$188 million strategic investment in September 2025, management later told CNA that the business generates enough cash each year to avoid another listing in the near term, and Straits Times reported management currently expects tariff costs to be absorbable. Those factors reduce immediate financing stress. They do not remove the need for evidence on commercial conversion and governance. The thesis should weaken materially if 2026-2027 approvals still fail to produce measurable APAC contribution, if manufacturing or tariff conditions stop being absorbable, if customer/distributor concentration remains opaque after diligence, or if the cloud-based surgical data platform cannot show enterprise-grade privacy, integration, and hospital adoption controls. [CR004, CR009, CR013, CR014, CR016]

7.5 Exhibits

8.1 Financing context and public pricing

UltraGreen does not look like a business that needs rescue capital. The assigned pool shows a roughly US$400 million IPO package in late 2025, a separate US$188 million strategic investment earlier that year, and an investor page still highlighting FY2025 revenue of US$142.4 million, adjusted EBITDA of US$89.4 million, and 85% gross margin. The harder judgment is valuation rather than survival. As of 16 October 2026, the investor page showed a US$0.475 last-done US-dollar line, which is materially below the US$1.45 IPO price and below every cited 2026 analyst target. That gap creates optical cheapness, but not yet a complete underwriting case, because the public record in this pool does not disclose an updated fully diluted share count, preference overhang, or refreshed FY2026 model.[CV001, CV003, CV004, CV005, CV006, CV010, CV016]

8.2 Thesis, anti-thesis, and what the market is debating

The long case is straightforward: UltraGreen still looks like a high-margin consumables franchise with enough evidence of software and platform optionality to justify some premium if execution holds. The prospectus and company materials frame PerfusionWorks and the UltraGreen Data Platform as a route from a dye-led business into a broader surgical-intelligence stack, while the refreshed regulatory evidence keeps the geographic expansion story alive. The anti-thesis is equally clear in the assigned pool. Singapore Business Review shows management preparing to respond to new US entrants based on pricing, distribution, and customer offer, while DBS argues the latest APAC approvals are groundwork rather than near-term earnings power because dye approvals still lag in most new territories. In other words, the stock may be cheap, but the next debate is about earnings quality and competitive durability, not awareness.[CV007, CV008, CV009, CV011, CV012, CV013, CV014, CV015]

8.3 Scenario framing, entry discipline, and final call

The public-market setup is attractive enough to keep UltraGreen on an active watchlist, but not clean enough to underwrite a full buy recommendation from this evidence set alone. A bull case exists if the current quote eventually re-rates back toward the historical analyst range once management proves that US competition is manageable and that the platform layer can add to, rather than merely decorate, the consumables business. The base case is still track: wait for refreshed FY2026 numbers, cap-table clarity, and evidence that APAC approvals become commercial rather than symbolic. The bear case is that the market is correctly discounting a business facing US pricing pressure, stale underwriting anchors, and a software story that remains earlier than the equity pitch implies.[CV001, CV003, CV007, CV008, CV009, CV011, CV012, CV013, CV016]

8.4 Exhibits