Harvey Startup Diligence Research Report
Harvey’s public case rests on legal-domain focus, workflow execution via agents, strong logo traction, and rapid geographic expansion. Confirmatory diligence should focus on whether growth quality, retention, margin profile, legal-data governance, and multi-model/vendor economics justify the pace of repricing and a potentially platform-level valuation.
Harvey Startup Diligence Research Report
Harvey screens as an active late-stage private legal-AI company with unusually strong public traction, deep blue-chip backing, visible enterprise controls, and a broad product stack. The biggest underwriting gaps remain financial quality, revenue concentration, cap-table detail, model/vendor dependence, and the unresolved public valuation mismatch between CB Insights ($8B) and March 2026 company/Wikipedia-linked $11B references.
No reviewed public source disclosed audited financials, revenue/ARR, burn, runway, debt, or margin structure, limiting underwriting quality.
Diligence request: Request audited and management financials, monthly metrics, debt schedules, and board packs.
Public valuation references diverge sharply between the current CB Insights unicorn list ($8B) and March 2026 company/Wikipedia-linked $11B signals, creating risk that the public narrative is stale, inconsistent, or structurally different from the actual financing documents.
Diligence request: Obtain signed financing docs, cap table, and a valuation bridge through the latest round and any secondary transactions.
Harvey’s logo traction is compelling, but public sources do not reveal whether a small set of flagship firms or enterprises drive a disproportionate share of ARR or usage.
Diligence request: Request top-customer revenue, cohort retention, usage expansion, logo concentration, and customer-reference calls.
Harvey’s product and GTM story depends on Azure, external model providers, retrieval partners, and workflow integrations such as Docusign and Datasite.
Diligence request: Review vendor concentration, contract terms, outage history, migration options, and gross-margin sensitivity to third-party cost changes.
Because Harvey handles privileged legal, diligence, and contract data, any gap between marketed controls and actual controls could be disproportionately damaging.
Diligence request: Validate SOC/ISO reports, DPA, incident logs, customer addenda, model-provider restrictions, and privilege-handling controls.
Incumbents such as Thomson Reuters and LexisNexis combine legal AI with proprietary content and existing distribution, while specialized rivals target drafting and contract workflows.
Diligence request: Review win/loss data, competitive displacement stories, renewal trends, and price pressure by segment.
Harvey’s public brand and customer traction are strong, but no quotas, CAC, pipeline-conversion, payback, or margin-by-segment data were found.
Diligence request: Request quota attainment, pipeline funnel, CAC payback, marketing spend, and sales-compensation metrics.
Rapid office expansion and multi-market hiring can create integration, quality-control, and compliance complexity before systems catch up.
Diligence request: Request headcount plans, regional P&Ls, implementation capacity metrics, and employee-retention data by office.
Public financing history is robust enough to reconstruct a multi-round trajectory, but audited financials, cap-table details, debt, taxes, and current valuation support remain private.
not publicly verifiable confidence: high
Public sources provide operating-scale proxies and workflow productivity anecdotes, but no audited statements, revenue, cash, backlog, or receivables aging were reviewed.
| metric | public signal | value | verification status | diligence request |
|---|---|---|---|---|
| Customer/account scale | Harvey cites both 1,300+ organizations / 100,000+ lawyers and 1,500+ customers in 60+ countries. | 1,300+ organizations; 100,000+ lawyers; 1,500+ customers; 60+ countries | partially_verified | Provide definitions for organization, customer, user, paid seat, and active user plus reconciliation over time. |
| Public revenue / ARR | No public revenue, ARR, cash burn, runway, margin, or EBITDA figures were found in reviewed sources. | not_publicly_verifiable | not_publicly_verifiable | Provide audited and management financial statements, ARR bridge, gross margin, burn, and runway. |
| Public pricing realization | Harvey reviewed pages route buyers to demos and do not publish list pricing. | not_publicly_verifiable | not_publicly_verifiable | Provide price book, discounting policy, seat utilization, implementation fees, and realized ASP by segment. |
| Workflow productivity / unit-economics proxy | PwC case study says teams can review 100 documents where they previously reviewed five; Vault page describes >80% faster review in an M&A transaction. | Case-study productivity only | partially_verified | Provide cohort-level ROI, deployment cost, support cost, gross margin, and model-inference cost per workflow. |
Public scale proxies are useful for screening but not sufficient for valuation underwriting.
not publicly verifiable confidence: medium
Growth direction is visible through financings and office expansion, but no public financial projections, hiring-plan budget, or financing assumptions were disclosed.
not publicly verifiable confidence: high
The investor roster is unusually visible for a private company, but share counts, dilution, employee ownership, debt, and off-balance-sheet obligations remain private.
| stakeholder group | public position | public evidence | diligence caveat |
|---|---|---|---|
| Sequoia | Lead/return investor across Series A, B, D, and March 2026 growth round | Named in Series A, Series B, Series D, Series E, and March 2026 financing announcements. | No share count, board rights, or pro-rata economics were disclosed publicly. |
| OpenAI Startup Fund and other AI/venture investors | OpenAI Startup Fund participated in A/B/C/D/E; GV, Kleiner Perkins, Coatue, Andreessen, GIC, DST, and others later joined. | Public round announcements repeatedly name OpenAI Startup Fund and other blue-chip investors. | Strategic rights, information rights, or model-distribution covenants are not public. |
| Strategic/legal-information investors | REV, the venture arm of RELX/LexisNexis owner, was named in Series D/E materials. | Series D and E announcements note REV participation and its RELX/LexisNexis parentage. | Any commercial distribution, content, or governance rights tied to REV are not publicly disclosed. |
| Founders, employees, option holders, debt holders | Not publicly broken out in reviewed sources. | No public cap table, debt schedule, option pool, or fully diluted share count was found. | Requires full cap table, option register, notes/SAFEs, and debt instruments. |
This table is directional only; no public shareholding percentages were reviewed.
partially verified confidence: medium
Financing history is public enough to build a credible round timeline, but tax positions, accounting policy detail, and the current valuation anchor require private confirmation and reconciliation.
| date | round or anchor | amount or signal | public valuation anchor | lead or participants | diligence request |
|---|---|---|---|---|---|
| 2023-04-26 | Series A | $21M | Not publicly disclosed in reviewed post | Sequoia; OpenAI Startup Fund; Conviction; SV Angel; Elad Gil | Obtain Series A term sheet, cap table, and post-money valuation. |
| 2023-12-19 | Series B | $80M | $715M valuation | Elad Gil; Kleiner Perkins; OpenAI Startup Fund; Sequoia | Confirm share class, dilution, and liquidation preferences. |
| 2024-07-23 | Series C | $100M | $1.5B valuation | GV; OpenAI; Kleiner Perkins; Sequoia; Elad Gil; SV Angel | Reconcile this company-announced valuation to later current-list anchors. |
| 2025-02-12 | Series D | $300M | $3B valuation | Sequoia; Coatue; Kleiner Perkins; OpenAI Startup Fund; GV; Conviction; Elad Gil; REV/RELX | Check any strategic-commercial rights linked to REV/RELX or other investors. |
| 2025-06-23 | Series E | $300M | $5B valuation | Kleiner Perkins; Coatue; Sequoia; GV; DST Global; Conviction; Elad Gil; OpenAI Startup Fund; Elemental; SV Angel; REV | Validate whether March 2026 financing created a new primary round or structured extension. |
| 2026-03-25 | Growth round / March 2026 valuation anchor | $200M | $11B valuation; more than $1B total funding | GIC; Sequoia; Andreessen Horowitz; Coatue; Conviction Partners; Elad Gil; Evantic; Kleiner Perkins | Obtain signed financing docs and reconcile to CB Insights current-list $8B anchor. |
| 2026-06-18 access date | CB Insights current unicorn-list anchor | Current-list row | $8B valuation anchor | CB Insights row also lists OpenAI Startup Fund, Kleiner Perkins, Sequoia Capital | Reconcile methodology and date basis for the current-list valuation anchor. |
CB Insights and company-announced valuation anchors conflict and should not be treated as interchangeable without primary documents.
Series A valuation was not disclosed in reviewed public materials.
Harvey’s public product story is rich and coherent, but pricing, SLAs, model economics, and module-level adoption are only partially visible.
partially verified confidence: medium
Harvey’s product footprint is unusually transparent for a private startup, but public materials do not resolve pricing, SLAs, model-routing economics, or module-level adoption.
| product | audience | key features | public evidence | verification status |
|---|---|---|---|---|
| Assistant | Lawyers, legal operations, professional-services teams | Source-grounded answers, deep research, drafting, document analysis, structured insight extraction. | Platform positioning and reviewed Assistant-page extracts. | verified |
| Vault | Diligence, investigations, compliance, transactional teams | Secure document storage, bulk analysis, compare data across thousands of documents, large vault limits. | Platform positioning and reviewed Vault-page extracts. | partially_verified |
| Knowledge | Research-heavy legal and regulatory workflows | Ground answers in legal databases, curated public sources, and institutional knowledge across jurisdictions. | Reviewed Knowledge-page extracts and platform listing. | verified |
| Agents / Workflow Agents | Practice groups and in-house teams automating repeatable legal tasks | Ad hoc planning, 500+ pre-built agents, and customer-specific agents that execute end-to-end legal work. | Agents page and March 2026 financing posts. | verified |
| Ecosystem / Integrations | Users working in Microsoft, DMS, email, mobile, and APIs | Word add-in, Outlook add-in, mobile apps, APIs/MCP, email workflow, partner connections. | Ecosystem page and partnership announcements. | verified |
| Contract Intelligence / Command Center / Shared Spaces | Enterprise legal teams and legal-ops leaders | Contract insights, analytics/benchmarking, and secure collaboration across organizations. | Platform overview listing and product-footprint language. | partially_verified |
Public product descriptions are strong, but availability, pricing, and SLA details vary by customer segment and are mostly private.
| offering | pricing signal | public terms | verification status | diligence request |
|---|---|---|---|---|
| Harvey | Request-demo / enterprise pricing | Reviewed Harvey product and company pages emphasize demos and sales contact rather than public list prices. | not_publicly_verifiable | Request seat-based, usage-based, and implementation pricing by segment plus discount policy. |
| Spellbook | Custom pricing based on number of team members; 7-day free trial | Spellbook pricing page says pricing is structured around the number of team members on a license and offers a 7-day trial. | verified | Benchmark Harvey against actual Spellbook proposals for comparable use cases. |
| CoCounsel Legal | Contact sales / no reviewed list price | Reviewed product page promotes demos and webinars; no public dollar-denominated list price was reviewed. | not_publicly_verifiable | Collect current CoCounsel enterprise proposals, add-on terms, and bundling assumptions. |
| Lexis+ with Protégé | Contact sales / no reviewed list price | Reviewed Lexis product pages describe workflows and content depth but do not disclose public list pricing. | not_publicly_verifiable | Obtain Lexis+ with Protégé commercial terms, bundle discounts, and content-access assumptions. |
Public pricing transparency in this category is limited; commercial diligence should rely on proposals and customer references.
Public sources do not provide a full low-level technical architecture.
Customer and partner evidence is broad and credible at the logo level, while revenue quality, concentration, and supplier economics remain largely private.
partially verified confidence: medium
Reviewed public sources show wide customer-logo breadth across law firms, professional services, and enterprise legal teams, but not ranked top-15 paid customers or spend by application.
| customer | segment | use case | public evidence | verification status |
|---|---|---|---|---|
| PwC UK Deals | Professional services / transaction advisory | End-to-end diligence, source-grounded analysis, and Vault-based document review. | Named customer case study quoting PwC partners. | partially_verified |
| Faegre Drinker | Law firm | Firmwide deployment after multi-phase evaluation. | Harvey deployment announcement. | partially_verified |
| Named in-house legal organizations | Enterprise legal departments | Advisory-board and customer-community participation at Dentsu, MongoDB, HSBC, NBCUniversal, Koch, Bridgewater. | In-house advisory-board announcement. | partially_verified |
| APAC legal teams and institutions | Law firms, public sector, and enterprise | Singapore-region customers and strategic-partner footprint including WongPartnership, Rajah & Tann, Drew & Napier, Attorney-General’s Chambers, Airwallex, and Singapore Judiciary relationship. | Singapore office announcement. | partially_verified |
| Italian law firms and enterprises | Law firms / in-house | BonelliErede firmwide adoption and named Italy footprint. | Milan expansion announcement. | partially_verified |
| Published customer-story roster | Law firms and enterprises | Customer-story titles on customers page include CMS, Burges Salmon, Dentsu, Reed Smith, Repsol, The Adecco Group, Syngenta, and others. | Harvey customers page. | partially_verified |
Public logos and case studies do not establish revenue concentration, depth of deployment, or net retention.
verified confidence: medium
Public partnerships with Docusign, Datasite, Mistral, Voyage, and regional institutions materially expand Harvey’s workflow reach and ecosystem story.
| partner | relationship | public evidence | gap or risk |
|---|---|---|---|
| OpenAI Startup Fund / OpenAI | Investor and model-development collaborator | OpenAI Startup Fund appears in multiple rounds; Series B notes collaboration with OpenAI on domain-specific foundation models. | Review commercial dependence, model-provider rights, data controls, and any exclusivity or MFN provisions. |
| Docusign | Strategic agreement-workflow integration partner | Joint announcement ties Harvey legal reasoning to Docusign IAM workflows and Iris assistant access to Harvey Knowledge. | Assess implementation complexity, usage economics, and customer adoption of combined workflow. |
| Datasite | Private-market / data-room workflow partner | Integration brings approved transaction materials into Harvey Assistant, Vault, Workflow Builder, and Word Add-In. | Validate security review, permission integrity, and GTM dependence on partner data access. |
| Mistral | Model provider / multi-model ecosystem partner | Mistral models are available in Harvey for eligible EU customers via early access. | Review model-routing policies, cost, fallback logic, and regional legal commitments. |
| Voyage AI | Embeddings and retrieval partner | Joint post describes a custom legal embeddings model trained on more than 20B tokens of case law. | Review IP, dependency, SLA, and replacement cost of the custom embedding layer. |
| Singapore Judiciary / PwC NewLaw / NUS Law | Regional ecosystem / education / institutional relationships | Singapore office announcement names a strategic partnership with the Singapore Judiciary, alliance with PwC NewLaw, and law-school partnership with NUS Law. | Determine whether these relationships convert into material revenue or durable channel advantage. |
Most relationship announcements do not disclose economics, exclusivity, or renewal mechanics.
not publicly verifiable confidence: high
Public sources disclose segment-level user and organization counts, not revenue by customer, so concentration remains a core diligence gap.
Public revenue concentration by customer is not disclosed.
inconclusive confidence: low
No significant severed customer, partner, or supplier relationships were confirmed in reviewed public sources, but the public-search evidence is weak and non-exhaustive.
partially verified confidence: medium
Supplier dependence is visible at the category level—Azure, model providers, embedding partners, and workflow integrations—but spend concentration and switch costs are private.
| supplier or dependency | role | public evidence | concentration or transition risk |
|---|---|---|---|
| Microsoft Azure | Cloud hosting and regional data processing substrate | Harvey says it hosts its cloud environment in Microsoft Azure with regional processing options. | Cloud concentration, regional availability, pricing, and shared-security assumptions should be reviewed. |
| External model providers (e.g., OpenAI, Mistral) | Foundation-model access within Harvey’s multi-model platform | Harvey says it is multi-model by design, requires zero data retention from model providers, and has Mistral available in early access for EU customers. | Model-routing, performance, cost, regional availability, and vendor leverage need confirmatory diligence. |
| Voyage AI | Custom legal embeddings and retrieval infrastructure | Voyage partnership post describes a Harvey-specific legal embeddings model. | Replacement cost, ownership, SLA, and exportability of the embedding layer are unclear publicly. |
| Datasite / Docusign / DMS connectors | Workflow and document-access integrations | Public integration announcements and ecosystem page emphasize embedded workflows in enterprise tools and data rooms. | Customer value may depend on maintaining partner APIs, permissions logic, and joint GTM cooperation. |
| Premium legal databases and curated public sources | Knowledge grounding and research authority | Knowledge and Assistant positioning emphasize premium legal databases, curated public sources, and trusted internal content. | Licensing scope, jurisdiction coverage, and cost of content access are not public. |
Public evidence reveals dependency categories, not their commercial or operational concentration.
Harvey appears differentiated on workflow execution and legal-domain focus, but it faces strong incumbent and specialist competitors with proprietary content or sharper point solutions.
partially verified confidence: medium
Harvey competes simultaneously against legal-information incumbents and focused workflow tools; public evidence supports strong execution depth but not a clean moat.
| competitor | segment | public funding or owner | product overlap | differentiator | source |
|---|---|---|---|---|---|
| CoCounsel Legal (Thomson Reuters) | Enterprise legal AI / research / drafting | Owned by Thomson Reuters | Research, analysis, drafting, and verifiable citation-backed workflows. | Incumbent legal-information distribution and workflow breadth. | Thomson Reuters product page |
| Lexis+ with Protégé (LexisNexis / RELX) | Legal AI research, drafting, and analysis | Owned by RELX / LexisNexis | Drafting, research, analysis, DMS integration, authoritative legal content, secure document analysis. | Deep proprietary Lexis content and DMS connectivity. | Lexis product page |
| Spellbook | Contract drafting and review for law firms and in-house teams | Private; pricing page does not disclose ownership details in reviewed extract | AI-assisted drafting, redlining, communications, and contract workflows. | Public free-trial and per-team-member commercial framing versus Harvey’s pure enterprise-sales posture. | Spellbook pricing page |
| General-purpose foundation models / generic AI tools | Horizontal AI assistants | Not comparable from reviewed sources | Drafting, summarization, and research assistance. | Harvey emphasizes legal-domain grounding, legal engineering, custom agents, and enterprise legal controls. | Harvey product, CTO, and security pages |
This matrix is directional and based only on reviewed public materials, not product testing or customer win/loss data.
| axis | harvey position | top competitor reference | public evidence | risk or implication |
|---|---|---|---|---|
| Legal-domain workflow execution | High | CoCounsel Legal; Lexis+ with Protégé | Harvey emphasizes agents, embedded legal engineering, and end-to-end workflow execution. | Advantage depends on maintaining customer-specific workflow depth and accuracy. |
| Authoritative content / source grounding | High but partner-dependent | LexisNexis and Thomson Reuters proprietary content stacks | Harvey positions around legal databases, curated sources, and institutional knowledge; competitors own major proprietary content libraries. | Incumbents may retain content/distribution advantages and pricing leverage. |
| Enterprise controls and regional deployment | High | Spellbook public security framing; incumbent enterprise suites | Harvey discloses Azure hosting, regional processing, annual audits, zero-training commitments, and broad legal documents. | Controls remain company-authored until validated against contracts and reports. |
| Pricing transparency | Low | Spellbook shows custom pricing with free trial; incumbents likewise skew enterprise-sales | Harvey does not publish list pricing on reviewed pages. | Opaque pricing makes public win/loss analysis difficult and may mask discount pressure. |
| Go-to-market distribution and installed base | Strong momentum, weaker legacy distribution than incumbents | Thomson Reuters and LexisNexis legal-information distribution networks | Harvey cites majority of AmLaw 100, 500+ in-house teams, and 60-country footprint, but incumbents still bundle within broader suites. | Harvey must keep converting workflow depth into durable renewal and expansion before incumbents catch up. |
Scores are qualitative analyst judgments anchored to public evidence, not customer win/loss datasets.
This is a qualitative positioning aid, not a market-share chart.
Harvey’s public GTM emphasizes direct enterprise selling, embedded enablement, and regional expansion, yet efficiency metrics and budget structure are not public.
partially verified confidence: medium
Harvey’s public GTM appears to combine direct enterprise selling, embedded legal engineering, ecosystem partnerships, and regional offices.
| channel or motion | public evidence | geographies or segments | implication or gap |
|---|---|---|---|
| Direct enterprise sales to law firms and in-house teams | Harvey cites majority of AmLaw 100, over 500 in-house legal teams, and 1,300+ organizations / 1,500+ customers. | Global law firms, corporate legal teams, asset managers | Need pipeline, CAC, sales-cycle, and renewal data by segment. |
| Embedded legal-engineering and implementation support | March 2026 funding posts emphasize embedded legal engineering teams; Singapore office spans legal engineering, customer success, sales, and G&A. | Top law firms, in-house legal departments, APAC expansion | Assess whether service intensity compresses gross margin or slows deployment capacity. |
| Partnership-led workflow distribution | Docusign and Datasite integrations place Harvey inside agreement and data-room workflows. | Contracting, M&A, private-market diligence | Joint-GTM economics and API/dependency risk should be validated. |
| International office-led expansion | Harvey announced offices/teams in Singapore and Milan and cites existing offices in London, Madrid, Munich, Paris, Bengaluru, and Sydney. | APAC and Europe | International scaling can stretch support quality and compliance operations. |
| Education / ecosystem development | Singapore post cites NUS Law partnership and law-school initiatives; company pages promote Harvey Academy. | Law schools, future talent pipeline, customer education | Measure whether education programs produce lower-cost adoption or mainly brand lift. |
The public GTM story is rich qualitatively but sparse on conversion, quota, and cost data.
| signal | date or period | public evidence | implication |
|---|---|---|---|
| Time100 Most Influential Companies mention | 2025 reference on company page | Company page says Harvey was named one of the Time100 Most Influential Companies. | Supports premium employer and brand positioning. |
| LinkedIn Top 50 Startup of 2025 mention | 2025 reference on company page | Company page says Harvey was named a LinkedIn Top 50 Startup of 2025. | Supports recruiting and awareness momentum. |
| CNBC Disruptor 50 mention | 2025 reference on company page | Company page says Harvey was on the Disruptor 50 list of the most innovative AI companies. | Reinforces category leadership narrative. |
| Newsroom cadence and expansion announcements | 2026 | Newsroom page shows dense cadence of funding, customer wins, partnerships, and office openings. | Suggests sustained PR and market-education activity. |
Brand-signal strength does not substitute for CAC, conversion, or efficient growth metrics.
not publicly verifiable confidence: medium
Major-customer logos and case studies are public, but relationship economics, expansion pace, and pipeline by major customer are not.
partially verified confidence: medium
Public evidence suggests Harvey wins by embedding into complex legal workflows and then expanding into custom agents, integrations, and regional support.
Qualitative funnel only; conversion metrics must be requested from management.
not publicly verifiable confidence: high
No public sales-compensation, average-quota, or sales-cycle model was found in reviewed sources.
not publicly verifiable confidence: medium
Public materials show significant brand activity and expansion, but not the budget capacity or cost structure behind the marketing plan.
Public R&D signals are above average for a private startup, with visible technical leadership, custom retrieval work, and a multi-model strategy, but cost and roadmap depth remain private.
partially verified confidence: medium
Public R&D evidence is stronger than average for a private startup, with a public CTO profile, legal-engineering emphasis, large document-scale claims, and visible security and platform work.
| name or role | role | background or public scope | source | verification status |
|---|---|---|---|---|
| Siva Gurumurthy | Chief Technology Officer | Previously CTO at Motive and Director of Engineering at Twitter; publicly outlines Harvey engineering priorities around platform, infrastructure, embeddings, and model quality. | Harvey CTO profile | verified |
| Winston Weinberg | CEO and co-founder | Quoted across financing and partnership announcements as CEO; Wikipedia unicorn list names him as a founder. | Harvey financing announcements and Wikipedia unicorn list | verified |
| Gabriel Pereyra | Co-founder | Named on Wikipedia unicorn list and credited in Harvey/Voyage technical post. | Wikipedia unicorn list; Voyage embeddings post | partially_verified |
| Harvey legal engineers (team) | Customer-embedded legal engineering | Financing and office-expansion announcements describe embedded legal-engineering teams and regional legal-engineering hiring. | March 2026 funding announcement; Singapore office announcement | partially_verified |
| In-house security team | Infrastructure, product, and operations security | Security page says Harvey’s in-house security team spans infrastructure, product, and operations with 24/7 coverage. | Harvey security page | partially_verified |
Public leadership visibility is incomplete; additional org-chart confirmation is needed for a full R&D diligence view.
partially verified confidence: medium
The public product pipeline includes partner integrations, model additions, custom embeddings, and scaled custom agents, but spend, attach rates, and commercialization timing are not fully public.
| initiative | status | expected or announced timing | source | verification |
|---|---|---|---|---|
| Custom agents / Agent Builder scale-out | Live and scaling | March 2026: 25,000+ custom agents on Harvey | March 2026 growth-round post | partially_verified |
| Docusign agreement-workflow integration | Announced | 2026-05-08 | Docusign partnership announcement | verified |
| Datasite live deal-data integration | Announced | 2026-06-09 | Datasite partnership announcement | verified |
| Mistral model family in Harvey | Early access for eligible EU customers | 2026-05-26 | Mistral partnership announcement | partially_verified |
| Voyage custom legal embeddings / firm-specific embeddings | Operational custom model plus additional roadmap | 2024-07-22 and ongoing | Voyage embeddings announcement | partially_verified |
Publicly announced launches do not confirm adoption, uptime, or unit economics.
Only a partial leadership and workforce picture is public; headcount, attrition, compensation, and full governance remain meaningful diligence gaps.
inconclusive confidence: medium
Only a partial public org chart could be reconstructed from reviewed materials; board composition and most operating leadership remain outside the public record reviewed here.
Public sources reviewed did not provide a full leadership roster or board structure.
inconclusive confidence: low
Public workforce disclosure is limited to proxy signals such as a 950+ associates metric, hiring in 16 markets, and region-specific team build-outs.
| signal | function or region | public evidence | diligence caveat |
|---|---|---|---|
| 950+ associates | Company-wide | Company page shows a 950+ associates metric on the careers/company page. | The company page does not define whether “associates” equals employees or another internal metric. |
| Hiring in 16 markets | Global | Company page says Harvey is currently hiring in 16 markets. | Open roles do not reveal actual filled headcount, attrition, or productivity by team. |
| APAC local team spans legal engineering, customer success, sales, and G&A | Singapore / APAC | Singapore office announcement describes functions and hiring areas. | No APAC team size or forecast was publicly disclosed. |
| Europe office footprint includes London, Madrid, Munich, Paris, and Milan | Europe | Milan expansion announcement lists European office footprint and ongoing local-team build-out. | Office footprint is not equivalent to headcount by function or location. |
Public workforce visibility is incomplete and relies on company-authored proxy metrics.
Exact employee headcount trend was not publicly disclosed in reviewed sources.
partially verified confidence: medium
A partial senior-management roster is public, but the full management bench and governance structure are not.
| name | role | public tenure or context | prior roles background | source |
|---|---|---|---|---|
| Winston Weinberg | CEO and co-founder | Quoted across financing and partnership announcements; named as founder in Wikipedia unicorn list. | not_publicly_verifiable in reviewed sources | Harvey March 2026 announcement; Wikipedia unicorn list |
| Gabriel Pereyra | Co-founder | Named on Wikipedia unicorn list and credited in Harvey/Voyage technical post. | not_publicly_verifiable in reviewed sources | Wikipedia unicorn list; Voyage technical post |
| Siva Gurumurthy | Chief Technology Officer | Interviewed as CTO on Harvey blog. | Previously CTO at Motive and Director of Engineering at Twitter. | Harvey CTO profile |
| John Haddock | Chief Business Officer | Quoted in the Singapore office announcement on APAC demand and expansion. | not_publicly_verifiable in reviewed sources | Singapore office announcement |
| Jorge Bestard | Vice President of EMEA Sales | Quoted in Milan expansion announcement on customer demand in Italy. | not_publicly_verifiable in reviewed sources | Milan expansion announcement |
This is not a full org chart or complete management roster.
not publicly verifiable confidence: high
No public key employment-agreement or benefit-plan detail was found in reviewed sources.
not publicly verifiable confidence: high
No public incentive stock-plan detail was found in reviewed sources.
inconclusive confidence: low
No significant employee-relations problems were identified in reviewed sources, but no comprehensive labor, NLRB, or court search was performed.
not publicly verifiable confidence: medium
Public turnover visibility is minimal; only rapid hiring and the lack of named departures in sampled public sources can be noted.
| subject | signal | date or period | source | diligence caveat |
|---|---|---|---|---|
| Named executives | No named executive departures were identified in the reviewed public company materials and sampled web searches. | As of 2026-06-18 review | Sampled company/newsroom pages and sampled search-result pages | Absence of public evidence is not proof of no departures; request HR and board records. |
| Company-wide workforce | Rapid hiring across 16 markets and new regional teams may mask underlying turnover dynamics. | 2026 public sources | Company page; Singapore and Milan announcements | Need rolling 24-month voluntary/involuntary attrition by function and location. |
| Retention and benefit plans | No public retention-plan or employee-benefit detail was found in reviewed sources. | As of 2026-06-18 review | Reviewed company and legal pages | Request equity-plan, bonus, retention, and severance documents. |
Turnover remains largely opaque in public sources.
Harvey publishes significant legal, privacy, and security materials, but litigation, IP registry coverage, insurance, and regulator history still require counsel-led confirmation.
inconclusive confidence: low
Sampled public searches did not surface a clearly matching lawsuit against Harvey, but this is not a clean legal opinion and should not be treated as dispositive.
| matter | court or forum | filed date | public status | source or search note |
|---|---|---|---|---|
| No clearly matching Harvey AI Corporation lawsuit surfaced in sampled top public search results reviewed in this run. | Sampled web search results | — | inconclusive | Top reviewed results were company pages and news rather than a clearly matching docket entry. |
| Alias query using “Counsel AI Corporation” did not produce a clearly matching lawsuit result in reviewed top results. | Sampled web search results | — | inconclusive | Top reviewed alias-query results were dictionary entries rather than a clearly matching docket entry. |
Counsel-led docket review is still required before treating litigation exposure as clean.
inconclusive confidence: low
Sampled public searches did not surface a clearly matching plaintiff-side case initiated by Harvey, but search coverage was limited.
| matter | court or forum | filed date | public status | source or search note |
|---|---|---|---|---|
| No clearly matching plaintiff-side case by Harvey AI Corporation surfaced in sampled top public search results reviewed in this run. | Sampled web search results | — | inconclusive | Sampled search coverage is too limited to conclude absence. |
| No clearly matching plaintiff-side case under the unsupported Counsel AI Corporation alias surfaced in reviewed top results. | Sampled web search results | — | inconclusive | Alias query did not surface a clearly matching docket; direct docket review remains required. |
This is not a substitute for a litigation schedule or outside-counsel letter.
not publicly verifiable confidence: low
As a software company, Harvey did not show public environmental or employee-safety issues in reviewed sources, but dedicated environmental and workplace-safety searches were not performed.
partially verified confidence: medium
Public sources support the Harvey AI Corporation brand and active model/embedding partnerships, but registry-level IP coverage and chain-of-title are incomplete.
| asset or right | public signal | jurisdiction or counterparty | status | diligence request |
|---|---|---|---|---|
| Harvey brand and site copyright | Reviewed legal/privacy pages use the Harvey AI Corporation name and include site copyright notices. | Company website / brand usage | publicly visible but not registry-verified | Confirm trademark filings, oppositions, and territorial coverage. |
| Custom legal embeddings and model IP | Voyage post describes a Harvey-specific embedding model and future firm/company-specific embeddings. | Voyage AI | public technical partnership; ownership and license rights not public | Review IP ownership, data rights, and assignment terms for custom embedding/model work. |
| Model and content licenses | Harvey legal page lists platform agreement, service terms, AI policy, and subprocessor/service-provider resources; security page references model-provider contractual restrictions. | External model providers, content partners, subprocessors | material but not publicly enumerated in detail | Review model-provider agreements, content-license schedules, and subprocessor contracts. |
| Patent/trademark registry coverage | Sampled public trademark query did not surface an obvious official filing in reviewed top results. | Trademark / patent registries | inconclusive | Run dedicated USPTO/EUIPO/WIPO and patent searches and obtain IP schedule. |
A complete IP diligence view requires registry, assignment, and open-source/license review.
not publicly verifiable confidence: high
No public insurance coverage detail was found in reviewed sources.
not publicly verifiable confidence: medium
Harvey publishes extensive standard-form legal documents and announces material commercial partnerships, but executed customer, vendor, and partner contracts remain private.
partially verified confidence: medium
Harvey publishes meaningful privacy, security, and legal disclosures, but no public enforcement actions were independently confirmed in this run.
| agency or framework | public signal | date | status | source |
|---|---|---|---|---|
| EU-U.S. / UK Extension / Swiss-U.S. Data Privacy Framework | Harvey privacy policy says Harvey has certified to the U.S. Department of Commerce under the DPF frameworks. | 2025-12-23 policy update | company-claimed compliance disclosure | Harvey privacy policy |
| GDPR / CCPA / cross-border transfer controls | Privacy policy and legal pages describe controller obligations, SCCs, privacy rights, and legal processing bases. | 2025-12-23 policy update; undated legal hub | company-claimed compliance disclosure | Harvey privacy policy and legal page |
| Security audit / standards references | Security page references annual SOC 2 Type II and ISO 27001 audits and commitments aligned with SOC 2, ISO, GDPR, and other standards. | — | company-claimed control environment | Harvey security page |
| Public enforcement actions or regulator proceedings | No specific enforcement action surfaced in reviewed company pages or sampled public search coverage. | — | inconclusive / requires counsel confirmation | Reviewed legal/privacy/security pages and sampled public search-result pages |
The absence of a reviewed public enforcement action is not a clean legal opinion.
No specific enforcement or litigation event was verified in reviewed sources.
| ID | Claim | Status | Sources |
|---|---|---|---|
| EC-001 | CB Insights current unicorn list shows Harvey with an $8B valuation anchor, U.S./San Francisco location, and investor names including OpenAI Startup Fund, Kleiner Perkins, and Sequoia Capital. | verified medium | SRC-027 |
| EC-002 | Wikipedia’s unicorn list records Harvey at $11B in March 2026, categorizes it under artificial intelligence in the United States, and names Winston Weinberg and Gabriel Pereyra as founders. | verified medium | SRC-026 |
| EC-003 | Harvey announced a $21M Series A led by Sequoia with participation from the OpenAI Startup Fund, Conviction, SV Angel, and Elad Gil. | verified medium | SRC-014 |
| EC-004 | Harvey announced an $80M Series B at a $715M valuation in December 2023. | verified medium | SRC-013 |
| EC-005 | Harvey announced a $100M Series C at a $1.5B valuation in July 2024. | verified medium | SRC-012 |
| EC-006 | Harvey announced a $300M Series D at a $3B valuation in February 2025, with participation from Coatue and REV/RELX in addition to returning investors. | verified medium | SRC-011 |
| EC-007 | Harvey announced a $300M Series E at a $5B valuation in June 2025. | verified medium | SRC-010 |
| EC-008 | Harvey’s March 2026 financing announcements describe a $200M round at an $11B valuation, more than $1B total funding, more than 100,000 lawyers using Harvey, the majority of the AmLaw 100 as customers, 500+ in-house legal teams, 50 asset-management firms, and 1,300+ organizations across 60 countries. | partially verified medium | SRC-008SRC-009 |
| EC-009 | Harvey’s company page describes the business as domain-specific AI for legal and professional services, says global law firms and Fortune 500 enterprises use Harvey, lists world-class investors, and says Harvey is used by 1,500+ customers in 60+ countries. | partially verified medium | SRC-001 |
| EC-010 | Harvey Agents are presented as a three-mode product with ad hoc, pre-built, and custom workflows, and Harvey says the library includes 500+ ready-to-use agents vetted by Harvey’s lawyers. | verified medium | SRC-002 |
| EC-011 | The Assistant product is positioned around source-grounded answers, multi-step analysis, deep research, and extracting structured insights from documents and trusted sources. | verified medium | SRC-006 |
| EC-012 | Vault is positioned as a secure document-analysis layer that can compare data across thousands of documents, store up to 100,000 files in a vault, and has customer-cited review-time savings. | partially verified medium | SRC-006 |
| EC-013 | Knowledge is positioned as a trusted-data layer that connects legal databases, curated public sources, and a customer’s institutional knowledge across jurisdictions. | verified medium | SRC-006 |
| EC-014 | Harvey’s ecosystem page shows integrations and access points including Word, Outlook, mobile, APIs/MCP, email workflows, and partner connections. | verified medium | SRC-006 |
| EC-015 | Harvey publicly states that it uses Microsoft Azure hosting, regional deployment options, contractual no-training commitments, annual SOC 2 Type II and ISO 27001 audits, and zero data retention requirements for model providers. | partially verified medium | SRC-003 |
| EC-016 | Harvey publicly maintains a legal hub with a platform agreement, service terms, support/SLA terms, security addendum, acceptable use policy, data-transfer addendum, DPA, privacy center, subprocessor list, law-enforcement requests page, and AI policy. | verified high | SRC-004 |
| EC-017 | Harvey’s privacy policy names the company as Harvey AI Corporation and states that Harvey has certified to the EU-U.S., UK Extension, and Swiss-U.S. Data Privacy Frameworks. | partially verified high | SRC-005 |
| EC-018 | Harvey’s customers page says the platform is trusted by over 1,000 leading law firms and enterprises and shows named customer-story titles including CMS, Burges Salmon, PwC/IFS, Dentsu, Reed Smith, Repsol, Deutsche Telekom/Gleiss Lutz, The Adecco Group, Syngenta, Munck Wilson Mandala, Cuatrecasas, and Gray Reed. | partially verified medium | SRC-015 |
| EC-019 | PwC UK’s Deals leaders publicly describe using Harvey end-to-end in diligence, source-grounded analysis, and Vault, with materially faster document interrogation and review. | partially verified medium | SRC-016 |
| EC-020 | Harvey states it now works with more than 500 in-house legal teams worldwide and publicly names inaugural advisory-board members from Dentsu, MongoDB, HSBC, NBCUniversal, Koch, and Bridgewater. | partially verified medium | SRC-017 |
| EC-021 | Harvey and Docusign announced a strategic partnership connecting Harvey’s legal reasoning platform with Docusign IAM workflows, including retrieval and analysis of specific agreements and workflow initiation from Harvey. | verified medium | SRC-018 |
| EC-022 | Harvey’s Datasite integration lets deal teams access approved transaction materials inside Harvey Assistant, Vault, Workflow Builder, and Word Add-In, while inheriting Datasite permissions. | verified medium | SRC-019 |
| EC-023 | Harvey’s Singapore expansion announcement names WongPartnership, Rajah & Tann, Drew & Napier, Braddell Brothers, Lee & Lee, Oon & Bazul, TSMP, Attorney-General’s Chambers, Sreenivasan Chambers, Airwallex, and a strategic partnership with the Singapore Judiciary, while describing APAC hiring across sales, legal engineering, and customer success. | partially verified medium | SRC-020 |
| EC-024 | Harvey’s Milan expansion announcement says BonelliErede is adopting Harvey firmwide, names multiple Italian customers, and says Europe now includes offices in London, Madrid, Munich, Paris, and Milan. | partially verified medium | SRC-021 |
| EC-025 | Faegre Drinker announced a firmwide Harvey deployment covering approximately 1,200 attorneys and consulting professionals after a multi-phase evaluation process. | partially verified medium | SRC-022 |
| EC-026 | Harvey’s CTO says the company is focused on platform, infrastructure, embeddings, integrations, and model quality, and describes Harvey as one of the world’s largest document-processing systems with enterprise customers uploading 200,000 to over 1 million documents. | partially verified medium | SRC-023 |
| EC-027 | Harvey and Voyage AI announced a custom legal embeddings model trained on more than 20 billion tokens of U.S. case-law text and said it reduced irrelevant top results by nearly 25% versus off-the-shelf models while using one-third the embedding dimensionality. | partially verified medium | SRC-024 |
| EC-028 | Harvey publicly describes itself as a multi-model platform and announced Mistral model availability in Harvey for eligible EU customers. | partially verified medium | SRC-025 |
| EC-029 | Thomson Reuters positions CoCounsel Legal as AI that unites research, analysis, and drafting in one seamless experience and emphasizes verifiable, citation-backed insights. | verified medium | SRC-028 |
| EC-030 | Lexis+ with Protégé is marketed as legal AI for drafting, research, and analysis grounded in LexisNexis content, web sources, and a customer’s DMS content. | verified medium | SRC-029 |
| EC-031 | Spellbook’s public pricing page discloses a custom, per-team-member pricing model with a 7-day free trial but no public list price. | verified medium | SRC-030 |
| EC-032 | Sampled public web searches for Harvey AI Corporation and Counsel AI Corporation litigation did not surface obvious docket entries in the top reviewed results, but this does not prove the absence of litigation. | inconclusive low | SRC-031SRC-032 |
| EC-033 | Reviewed public legal pages identify Harvey AI Corporation rather than Counsel AI Corporation as the corporate name. | verified high | SRC-005 |
| EC-034 | Harvey’s company page presents a 950+ associates metric and says the company is currently hiring in 16 markets, but the term “associates” was not independently corroborated as employee headcount. | inconclusive low | SRC-001 |
| EC-035 | Sampled public trademark search queries did not surface an obvious official Harvey AI filing in reviewed top results, so Harvey’s trademark schedule remains incomplete in public evidence reviewed here. | inconclusive low | SRC-033 |
This report is a public-evidence diligence snapshot, not investment advice. Important financial, legal, technical, and contractual facts remain non-public and should be verified directly with management and primary documents before any investment decision.