Zhisquare

Mixes official company disclosures with third-party financing and order reports; undisclosed metrics are explicitly preserved as gaps.

The public leadership surface is founder-heavy; the broader executive bench and board structure are not fully disclosed.

Dates use public announcement dates rather than internal board-approval dates; the table is the single chronology of record.

The syndicate is unusually broad; the central open question is how much of its value is strategic access versus pure financial underwriting.

Boundary rows are analytical definitions based on cited market and company materials; they are not an industry-standard taxonomy.

This table mixes analyst estimates and policy targets; rows are not additive and should not be combined into one TAM.

Buyer, user, and payer roles are inferred from policy deployment logic and named scenario evidence rather than disclosed contract language.

Drivers and constraints blend observed market facts with explicit diligence asks where public ROI proof is still limited.

Rows compare public signals only; several private competitors still lack disclosed pricing, customer counts, or unit economics.

Cells are ordinal and evidence-backed rather than benchmark scores; unknown public evidence is treated conservatively.

Only Unitree publishes a list price in retained sources; all other rows intentionally preserve pricing opacity rather than guessing.

This register focuses on moat durability, not general operating risk; several mitigations require private diligence rather than more web research.

Most rows are intentionally null-status because public sources discuss mechanisms and use cases more clearly than booked revenue.

Absence of public price is analytically meaningful because it blocks normal unit-economics triangulation.

Headline fundraising is strong, but adequacy cannot be underwritten without cash, burn, and commitments behind the scale-up plan.

Public evidence supports the importance of these metrics but not their values; null is the correct answer.

This table deliberately records the missing finance package that would be standard in serious underwriting.

Rows combine official product pieces with the public workflow layer around them; maturity is based on evidence visibility, not internal readiness scores.

This table is architecture logic extracted from company and technical surfaces, not a full engineering bill of materials.

Benefits reflect company or media framing; the public record still lacks standardized KPI readouts by workflow.

The roadmap records public product and research milestones, not internal sprint or release cadences.

Public compliance evidence is thinner than architecture evidence; that imbalance matters for real deployment risk.

Rows distinguish industrial anchors from adjacency segments; the strongest public proof remains concentrated in manufacturing.

Trajectory evidence is strongest on one industrial anchor and one retail-operations claim; broad fleet-base disclosure remains absent.

Named proof mixes hard orders, strategic partnerships, and planned rollouts; investors should not treat all rows as equal-quality evidence.

Nulls are the correct public answer on retention; the chapter records what must be requested rather than guessing.

This table focuses on growth quality, not whether demand exists at all; Zhisquare clearly has demand signals, but concentration could distort interpretation.

The table covers the surfaced regulatory and legal exposures visible in retained public sources; it is exhaustive only for the public surface reviewed here.

Operational risk is driven more by under-disclosed reliability and support systems than by lack of product ambition.

Dependency risk is not just supplier concentration; it also includes capital, procurement, and flagship-customer dependencies.

People risk centers on whether the company can professionalize as fast as it scales.

These kill criteria translate broad risks into measurable decision rules.

The thesis table is framed around what evidence would move the decision, not around static opinions.

Comparable set is model-appropriate rather than multiple-pure; each row answers a different valuation question.

These scenarios are qualitative because the public record is too thin for honest DCF-style precision.

Kill triggers are designed for investment-committee monitoring, not just post-mortem explanation.

Recommendation is explicitly price-sensitive; it is not a company-quality rejection.

These asks define the minimum package needed to convert today's story into an investable pricing decision.