Model ML: Initiation of Coverage

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.

Model ML competes most directly with Rogo for banking deliverables and with Endex for Office-native output, with Hebbia and frontier-lab finance products as broader alternatives. Its consulting and accounting client base — PwC, Deloitte, EY by company account — is unusual in our coverage.

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

• Tier-one deployments: HSBC as customer and investor, and company-announced rollouts at PwC, Deloitte's M&A practice and FT Partners.
• Growth and capital: From a $12m seed to over $100m raised in about eighteen months, with reported talks at a valuation above $1bn and a former Trade Desk CFO hired.
• Deliverable focus: Decks, memos and reports in the firm's house template, document review (AutoCheck) and Word, Excel, PowerPoint and Teams add-ins.
• Deployment flexibility: Single-tenant environments, deployment into the customer's own Azure, open-source models in the client environment and customer-managed keys.
• Model-agnostic engineering: Routing across frontier and open models, a published Composite methodology and a fine-tuned slide checker run at much lower cost.
• Data ecosystem: Capital IQ, FactSet, PitchBook, LSEG, Preqin, Daloopa, Third Bridge, Quartr and data-room connectors, plus customer-built MCP servers.
• Shipping cadence: Monthly product updates and new models supported within days of release.

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 Model ML. 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.

• Attestation consistency: The homepage claims SOC 2 Type II, while the security page lists ISO 27001 only; no report period or auditor was located. Buyers should request the report.
• Governance controls: SCIM, zero data retention with model providers, MNPI and information-barrier controls and a named subprocessor list were not located.
• Performance evidence: Benchmarks are company-run; results published by OpenAI show 83% of key Excel outputs correct but only half of workbooks fully correct. We encourage a bake-off on the firm's own templates.
• Related-party testimonials: FT Partners is lead investor, transaction adviser and named customer at once.
• Data licences: PitchBook and LSEG require the client's own licence; total cost should include data.
• Inconsistent figures: The Big Four count, founding year and office count differ across sources, and no audited ARR has been disclosed.
• Channel governance: Access through web, Office, Teams, Slack and other channels widens the governance surface.
• Website instructions to AI agents: Company web pages contain hidden text addressed to AI agents; firms using AI tools for vendor diligence should be aware of this.

Model ML was founded in 2023 by brothers Chaz Englander, who previously founded Fat Llama, and Arnie Englander, who founded Fancy — both earlier Y Combinator companies that were acquired. It went through Y Combinator's winter 2024 batch, began as software for the founders' family office, and operates from New York, San Francisco, London, Hong Kong and Madrid, with engineering in London.

Model ML emerged from stealth with a $12m seed in 2025 and raised a $75m Series A led by FT Partners in November 2025, with Y Combinator, QED, 13Books, Latitude and LocalGlobe. HSBC Asset Management made a strategic investment in August 2026, when the company said it had raised over $100m; talks for a further round above a $1bn valuation were reported in September. Alex Kayyal, formerly CFO of The Trade Desk, became CFO in October 2026.

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

Model ML presents an agentic workspace organised around deliverables. Figure 2 sets out our assessment of how it maps onto the investment workflow.

Agent and Workflows

Goal-driven and multi-agent runs that produce decks, reports and memos in the firm's template.

Grids and Document Review

Grid analysis across documents and AutoCheck review of presentations, from the Flippr acquisition.

Office, Teams and Slack

Word, Excel, PowerPoint and Teams add-ins (July 2026) and a Slack agent (September 2026).

Notetaker, Sites and dashboards

Meeting notes, shareable sites and agentic dashboards.

Scope of the product today

We did not locate broker research, public API documentation or native Excel model building from scratch.

Model ML routes each task to the best-suited model across OpenAI, Anthropic and open-source families, and can run open models in the client's environment. The platform runs on Azure with per-customer databases and audit trails in Azure Monitor.

Its Composite benchmark publishes a methodology with deterministic checks and cross-model judging, but results are company-run. OpenAI's case study reports 83% of key Excel outputs correct and half of workbooks fully correct.

Licensed and partner data include Capital IQ, FactSet, PitchBook, LSEG, Preqin, Daloopa, Dun & Bradstreet, Quartr, Bigdata.com and Third Bridge, alongside a global filings connector. Client content arrives through email, storage, Salesforce, Snowflake and Intralinks and Blackroom data rooms.

PitchBook and LSEG require the client's own licence; the licensing basis of other sources was not stated. Broker research was not located.

Model ML fits the production end of banking, advisory and consulting work: pitch books, diligence reports and IC memos in the firm's template, reviewed by AutoCheck.

For buy-side readers the fit is narrower: monitoring arrived as dashboards in August 2026, and modelling centres on updating existing workbooks.

Model ML offers single-tenant and customer-Azure deployment, customer-managed keys, SAML/OIDC SSO with JIT provisioning, TLS 1.2 and AES-256 encryption, and a commitment never to train on customer data, with ISO 27001 and a Vanta trust centre. Figure 3 sets out what our public review located.

We would encourage buyers to request the SOC 2 report, SCIM availability, zero-data-retention terms, information-barrier controls and the subprocessor list.

The offsetting factors are real. Single-tenant deployment in the client's own cloud is among the strongest deployment options in our coverage, and HSBC chose to invest after deploying.

Model ML does not publish pricing; enterprise plans are scoped per engagement and no free trial is mentioned.

The value case is direct substitution of analyst and consultant hours on deliverables. Buyers should include data licences and governance requirements in the total cost.

Model ML sits in the Innovators quadrant of the GAII Intelligence Matrix, with market strength above the line on momentum and AI, and investment utility held back by buy-side fit and data.

GAII initiation universe, 20 October 2026. Scores as currently published on the GAII Intelligence page; preliminary and based on public information.

11.1 Where Model ML is strongest relative to peers

• Against research platforms, on the finished deliverable in the firm's own template.
• Against Excel copilots, on breadth across Word, PowerPoint, Teams and Slack.
• Against SaaS-only vendors, on single-tenant deployment in the client's cloud.

11.2 Areas of narrower functional focus

Model ML has built depth in deliverables and deployment rather than in buy-side research or governance documentation.

• Governance: SOC 2 position inconsistent; SCIM, ZDR and barriers not located.
• Content: no broker research.
• Buy-side: limited monitoring and idea generation.
• Evidence: company-run benchmarks; no audited ARR.

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 report, type, period and auditor.
• Availability of SCIM.
• Zero-data-retention terms with model providers.
• Information-barrier and MNPI controls.
• The subprocessor list.
• Named data-residency regions.
• Absolute ARR and seat counts.
• Which data sources are included and which need client licences.
• Pricing and contract terms.
• Status of the reported funding round.
• Plans for broker research.
• Accuracy evidence on Excel output.

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 Model ML specifically.

If you are a fundamental equity manager

• Run a bake-off on your own templates, especially Excel.
• Test AutoCheck on a live deck.
• Ask for references beyond the investor base.

If you are responsible for technology or vendor risk

• Open with Figure 3; request the SOC 2 report.
• Consider single-tenant deployment in your Azure.
• Set governance across Office, Teams and Slack channels.

If you are assessing the category rather than the company

• Deliverable agents are where adoption is fastest.
• Single-tenant deployment is becoming the enterprise standard.
• Hidden instructions on vendor sites are a new diligence risk.

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