Fintool

Fintool was an AI-native financial research platform built specifically for institutional investors — hedge funds, asset managers, and investment banks. Founded in 2023 and backed by Y Combinator, Menlo Ventures, and Cassius, the company built proprietary AI agents for qualitative financial analysis of SEC filings, earnings calls, and transcripts. It was acquired by Microsoft in April 2026; the platform is no longer sold as a standalone product and its technology is being integrated into Microsoft 365 / Office for financial services. As of June 2026, fintool.com redirects to the founder's acquisition announcement. Firms evaluating this vendor should treat it as a discontinued standalone tool.

Natural language chat interface over SEC filings (10-K, 10-Q, 8-K, proxy statements), earnings call transcripts, and investor day presentations with source citations on every answer. Company screener for identifying industry trends and investment themes across earnings calls at portfolio scale. Real-time AI feed monitoring new 8-K filings and significant company events. Automated earnings template generation — upload a template, Fintool populates it for the most recent quarter. Portfolio-level intelligent alerts (e.g., notify when key themes shift across trailing earnings calls). Comprehensive equity research reports covering business model, financials, and competitive dynamics. Internal data integration (Enterprise): upload proprietary firm data for combined analysis with public financials. Consensus estimates integration (v4+) via S&P Global covering revenue, EPS, and price targets from sell-side analysts.

Purpose-built AI agent, not a wrapper on general-purpose LLMs. Fintool-v4 scored 63% better than the next-best agent on the Finance Agent Benchmark (developed by vals.ai with Stanford researchers) while running 4x faster and at 10x lower cost. Internally benchmarked at ~90% accuracy on financial question answering tasks. Fintool-v5 (January 2026) introduced fully agentic workflows: the agent autonomously builds DCF models in Excel, earnings decks in PowerPoint, and research memos in Word without step-by-step user prompting. REST API available (Enterprise tier only): endpoints at api.fintool.com/v2/chat (natural language financial queries with citations) and api.fintool.com/v1/search (ranked document chunk retrieval for RAG pipelines); supports real-time streaming via Server-Sent Events; Bearer token authentication. No MCP server was publicly documented or listed for Fintool prior to acquisition. Post-acquisition, API access and ongoing development roadmap are subject to Microsoft's integration decisions and not publicly confirmed.

Two tiers were offered prior to acquisition: a Regular plan for small-to-medium firms and an Enterprise plan with API access, internal data integration, and dedicated cloud environments. Specific per-seat or per-month pricing was not publicly listed; Enterprise required contacting sales. As of April 2026 acquisition by Microsoft, the standalone product is no longer sold. Future access to Fintool capabilities is expected via Microsoft 365 / Copilot subscriptions.

Microsoft Excel (output generation, DCF model population), Microsoft PowerPoint (earnings deck generation), Microsoft Word (research memo generation), Microsoft OneDrive. S&P Global consensus estimates (v4+). Cloud provider integrations for Enterprise internal data upload (specific providers not publicly named). No native CRM, Slack, or third-party workflow integrations were documented. Post-acquisition, integration roadmap falls under Microsoft 365 / Agent 365 ecosystem.

Enterprise-grade security advertised: end-to-end encryption, compliant data storage, and dedicated cloud environments for Enterprise customers. Isolated data storage per customer noted in product materials. SOC 2, ISO 27001, or other formal certification status was not publicly confirmed on a dedicated trust/security page prior to acquisition. Post-acquisition, security posture defaults to Microsoft's enterprise compliance framework (SOC 2, ISO 27001, FedRAMP, etc.) as capabilities migrate into M365.

Founded 2023 in San Francisco by Nicolas Bustamante (CEO) and Edouard Godfrey (CTO). Y Combinator Winter 2023 batch. Raised approximately $7.24M total across multiple rounds (per PitchBook); investors included Y Combinator, Menlo Ventures, Cassius (Jean de La Rochebrochard and Emmanuel Seuge), and angel investors including founders of Datadog and HuggingFace. Team peaked at ~6 employees. Reached $440K ARR by September 2025. Notable customers included Artisan Partners, First Manhattan, PwC, and UBS. Acquired by Microsoft in April 2026; the six-person team joined Microsoft's Office Product Group under Sumit Chauhan. Acquisition terms not publicly disclosed.

Moderate. Fintool's core capability — autonomous qualitative analysis of public company filings, earnings calls, and transcripts — is directly applicable to a firm's sector research workflow (analyzing public comps, tracking earnings themes in deep-tech sectors, Energy, etc.). The Finance Agent Benchmark performance and low cost-per-query were genuine differentiators for high-volume research tasks. However, as a standalone product Fintool is discontinued following the April 2026 Microsoft acquisition. Fit is now contingent on Microsoft delivering the promised M365/Copilot integration for financial services, which has no confirmed timeline. Firms needing this capability today should evaluate AlphaSense or build equivalent workflows via the Fintool API (if still accessible under Microsoft) before it is formally sunset.

Discontinued as a standalone product as of April 2026 — acquired by Microsoft and no longer sold independently. Future availability, pricing, and feature roadmap are entirely subject to Microsoft's M365 integration timeline, which is not publicly confirmed. Prior to acquisition: coverage limited to public company filings and transcripts (no private company data, no proprietary deal data enrichment). English only. No confirmed SOC 2 or ISO 27001 certification. API access gated to Enterprise tier with sales-led access. No MCP server documented. Small team (6 people) with limited enterprise support infrastructure.