Endex: Coverage Update — 21 Oct 2026

GAII Intelligence defines the AI Research & Intelligence market as software that applies generative and agentic AI to the research process of professional investors: locating candidate securities, assembling and interrogating evidence, producing written analysis with traceable sources, and monitoring positions once held. Buyers are hedge funds, long-only managers, multi-manager platforms, family offices and allocators; the economic buyer is typically the chief technology officer, the head of research, or a portfolio manager with a technology budget.

Capabilities we treat as mandatory

• Evidence retrieval at scale: the ability to search and reason across regulatory filings, transcripts and news, and across documents the client supplies.
• Source traceability: every material claim in an output must be reachable back to the passage that supports it.
• Institutional data handling: a stated position on where client data goes, how long it is held, and whether it is used for model training.

Capabilities we treat as differentiating

• Persistent state: the system remembers a thesis, a coverage list or a position between sessions and acts on it without being re-prompted.
• Workflow surfaces: delivery inside the applications analysts already use — the spreadsheet, the research management system, email and messaging.
• Owned or exclusive content: material that a competitor cannot obtain by signing the same third-party contract.
• Published evaluation: a measured, disclosed accuracy methodology rather than an assertion of quality.

Endex competes at the production end of the research workflow — models, memos and decks — rather than in research discovery. Its closest comparisons in our coverage are Daloopa for Excel-native fundamentals and Rogo for banking deliverables. Outside our coverage, Microsoft Copilot Agent Mode in Excel, Claude in Excel and Shortcut are direct alternatives for the Excel workflow.

The following are the capabilities and characteristics that, in our assessment, most distinguish Endex within the AI Research & Intelligence market. Each is drawn from publicly documented product capability, published interviews with management, or independent hands-on review.

• Office-native delivery: Agents work inside Excel, PowerPoint, Word and Outlook, with delegation by email, Teams or Slack, plus a web workspace, mobile app and MCP/API. No other vendor in our coverage reaches as many of the surfaces analysts deliver in.
• Deliverable fidelity: Slides are built and restyled to match the target deck, and models follow firm templates. The output arrives in the firm's own file rather than a separate workspace.
• Model auditing: Cell-reference tracing, reconciliation across tabs and detection of broken references, circular logic and hard-coded totals. A hands-on review showed these working on a real workbook.
• Connector breadth: Data rooms including Intralinks, SharePoint, Box and Dropbox, DealCloud and Affinity, ERP and warehouse sources, real-estate systems, and filings across the US, UK, Canada, Japan, Hong Kong and Australia.
• Lab backing: The OpenAI Startup Fund led the $14m seed, and OpenAI has published a customer story describing reinforcement fine-tuning work with Endex.
• Recognisable logos: Hg, TA Associates, Centerbridge, Moelis, Raine, RXR and Colliers are among fifteen firms marketed as enterprise deployments.
• Shipping velocity: The product expanded from one Excel agent to a full office suite in about fourteen months.

The following are factors that prospective clients should weigh, and areas where our public review did not find sufficient evidence to reach a settled view. Several are common across vendors at this stage of the market and are not specific to Endex. We encourage prospective clients to raise each of them directly with the company; we have done the same, and will update this note as answers arrive.

• Customer evidence: The fifteen marketed firms are presented as logos; we did not locate a customer quote, case study or press release for any of them, and G2 shows no reviews. Prospective clients should ask for references.
• Independent accuracy evidence: Wall Street Prep was unable to test Endex for its 2026 modelling ranking, and the published benchmarks are vendor and partner figures from the o1 era. A sponsored review found a margin logic error in a generated DCF. We encourage a bake-off against Microsoft Copilot Agent Mode, Claude in Excel and Shortcut.
• Certification wording: The security page shows SOC 2 Type II and ISO 27001 and refers to independent audits; a November 2025 company post described ISO-aligned controls. The trust centre was not publicly readable. Prospective clients should request the reports and the ISO certificate.
• Data retention: Company materials describe client data as never stored, while the privacy policy says some data is retained as part of task records and the security page describes logging of every interaction. Product data terms sit in non-public agreements and should be confirmed in contract.
• Model routing: The integrations page lists eleven model providers, including several Chinese-origin models, and requests route through Cloudflare AI Gateway. Buyers should confirm which providers process their data and whether models can be restricted.
• Enterprise controls: SSO/SAML, SCIM, single-tenant or VPC deployment, regional hosting and an information-barrier position were not located.
• Scale and focus: About nineteen staff and $14m raised, with no round located since August 2025. Positioning has broadened from a finance analyst to a general office suite; buyers wanting finance depth should probe roadmap focus.
• Commercial access: Pricing is not published and access was waitlist-gated into 2026. Premium data feeds such as FactSet and Capital IQ appear to rely on the client's own entitlements.

Edvise, Inc., trading as Endex, was founded in 2022 according to third-party databases. Co-founder and chief executive Tarun Amasa left Duke University as a Thiel Fellow; co-founder Pratham Soni is named in OpenAI's customer story. The team worked from OpenAI's San Francisco office during 2024–25 and is now headquartered in New York.

Endex announced a $14m seed round led by the OpenAI Startup Fund in August 2025, with Neo and executives from Palantir, Two Sigma and Coatue participating. No later round has been located. A company-filled job board lists about nineteen employees and eight open New York roles, mostly in go-to-market and operations. Customer counts and revenue have not been disclosed.

Company-reported figures are shown as such and are not independently verified. Verification is carried into the vendor questionnaire.

Endex now presents itself as an AI office suite: agents that work inside the Microsoft Office applications, assigned and reviewed from a web workspace. Figure 2 sets out our assessment of how the product maps onto the investment workflow.

Excel

Builds, updates and audits models; traces cell references; reconciles across tabs; converts PDFs to tables; pulls filings into the model.

PowerPoint and Word

Builds and restyles slides to the target deck; drafts and revises memos with claims traced to sources.

Outlook, Teams and Slack

Tasks assigned by tagging Endex in an email or chat; results delivered back into the thread.

Web workspace, mobile and MCP/API

Assign, track and review agents; trigger agents from the firm's own tools. API documentation was not located.

Scope of the product today

We did not locate persistent monitoring, a public-equity screening surface, Google Sheets support or published API documentation. Launch dates for the PowerPoint, Word, Outlook, mobile and MCP products were not located.

Endex was built on OpenAI reasoning models. OpenAI's customer story describes reinforcement fine-tuning on GPT-4o mini and o1-series models, and agents that retrieve and reflect with reasoning models rather than a standard retrieval pipeline. The integrations page now lists eleven model providers, and requests route through Cloudflare AI Gateway.

Evaluation evidence is the area to probe. The published figures — a 70% blind-test win rate and 99% multimodal extraction accuracy for o1 — are vendor and partner benchmarks without published method, and no independent test was located.

Listed sources include SEC filings, earnings calls, and filings from Companies House, SEDAR+, EDINET, HKEX and ASX, with FactSet, Capital IQ and Morningstar as integrations. Client content reaches the agents through data rooms, document stores, CRM, ERP and warehouse connectors, and real-estate systems such as Yardi, CoStar and Argus.

The site does not state whether premium market data is licensed by Endex or requires the client's own entitlement; a third-party review describes the product as relying on user-provided data. Expert-network and broker research were not located.

Endex earns its Workflow score on modelling and delivery. Building, updating and auditing models inside Excel, and producing decks and memos in the firm's own templates, is where analysts spend production time, and no peer reaches as many Office surfaces.

The narrower areas are upstream. We did not locate idea generation for public equities or persistent thesis monitoring, and the marketed client base leans to private equity, advisory and real estate rather than public-markets research.

Endex presents more security and privacy detail than many seed-stage peers: SOC 2 Type II and ISO 27001 badges, AES-256 encryption, AWS KMS key management, RBAC with enforced MFA, logging of every interaction and a no-training-by-default position. Figure 3 sets out what our public review located.

We would encourage any firm evaluating Endex to request the SOC 2 Type II report and ISO 27001 certificate, the subprocessor list and model-provider routing, product-data retention terms, and the availability of SSO/SAML, SCIM and single-tenant deployment.

The offsetting factors are real. The Vanta-monitored control set, US data storage with standard contractual clauses and zero-data-retention language with model providers are a stronger starting position than most companies at this stage publish.

Endex does not publish pricing. A founder was reported in 2025 as expecting banks and consulting firms to pay at least $1,000 per employee, without a stated period, and a sponsored creator offer referred to a first-year discount, which suggests annual subscriptions.

The value argument is direct analyst-hour substitution on modelling and deck production. Buyers should negotiate a pilot with defined success criteria and confirm whether data feeds require their own licences.

Endex sits in the Emerging quadrant of the GAII Intelligence Matrix, with Investment Utility of 6.48 held back by upstream workflow coverage and content, and Market & Innovation Strength of 6.02 by enterprise and evaluation evidence. Peer positions shown are as currently published; the remaining tranche-1 companies are being recalibrated to the same standard.

GAII coverage universe, 21 October 2026. Peer scores as currently published on the GAII Intelligence page; the remaining tranche-1 companies are being recalibrated to the full sub-criteria standard. Preliminary and based on public information.

11.1 Where Endex is strongest relative to peers

• Against research platforms, on delivery: the model, memo or deck arrives in the firm's own Office file.
• Against Excel-only tools, on breadth of surfaces — PowerPoint, Word, Outlook, Teams, Slack and MCP alongside Excel.
• Against general-purpose copilots, on finance-specific auditing and connectors to data rooms and deal systems.

11.2 Areas of narrower functional focus

Endex has built breadth across production surfaces rather than depth in research content or monitoring.

• Content: no expert-network or broker research; premium feeds appear to rely on client licences.
• Research workflow: no persistent monitoring or public-equity screening located.
• Evidence: no named customer statement or independent accuracy test located.
• Enterprise: trust centre not publicly readable; SSO and deployment options not located.

Scores are on a 0–10 scale. Each dimension score is the mean of its sub-criteria, and the GAII Overall Score is the weighted mean of the six dimensions. The detailed sub-criteria scores that build each dimension are set out in Appendix A.

These are the areas where our public review did not find sufficient evidence to form a settled view. We have put them to the company and will update this note when they are answered. Prospective clients may find them a useful agenda for their own evaluation call.

• The SOC 2 Type II report period and auditor, and the ISO 27001 certificate and certifying body.
• Access to the trust centre and the subprocessor list.
• Which model providers process client data, whether zero data retention applies to each, and whether models can be restricted.
• Product-data retention and deletion terms.
• Availability of SSO/SAML, SCIM and single-tenant or VPC deployment.
• Whether FactSet, Capital IQ and Morningstar access requires the client's own licence.
• Which of the fifteen marketed firms are paying enterprise deployments, and two references.
• Pricing, contract term and pilot structure.
• Whether capital has been raised since August 2025.
• Headcount, engineering share and the leadership team beyond the founders.
• Results of any independent accuracy evaluation, and willingness to participate in one.
• Roadmap focus between finance-specific depth and the general office suite.

This note is an assessment of a vendor against a defined framework, not a recommendation to buy or to decline. The following are the considerations we would put at the centre of an evaluation of Endex specifically.

If you are a fundamental equity manager

• Run a bake-off on your own models against Copilot Agent Mode, Claude in Excel and Shortcut.
• Test deck restyling against your house template and judge how much rework remains.
• Ask for two references from firms like yours among the marketed logos.

If you are responsible for technology or vendor risk

• Open with Figure 3; request the SOC 2 report, ISO certificate and trust-centre access.
• Confirm model-provider routing and restrict providers where required.
• Settle product-data retention and SSO/SCIM before any firmwide deployment.

If you are assessing the category rather than the company

• Office-native agents are moving the competitive line from research to the finished deliverable.
• Microsoft's own Agent Mode raises the bar for specialist Excel vendors on accuracy and integration.
• Independent accuracy testing is becoming the expected evidence for modelling tools.

Each dimension score in Section 12 is the mean of the sub-criteria below.