Formula Insight

Based in New York, Formula Insight develops a centralized platform for institutional investors to manage, track, and query their Excel financial models. The software enables conversational searches across millions of spreadsheet cells and cross-references key performance indicators with SEC filings, earnings transcripts, and research papers. To further support equity research workflows, the system integrates directly with FactSet data, allowing users to plot their internal projections against consensus estimates. Operating with a team of two employees, the enterprise software startup targets hedge funds managing billions in assets and is backed by Y Combinator as part of its Summer 2024 accelerator cohort. Following its initial development phase, the company launched its new equity research platform in December 2024. Formula Insight was founded in 2024 by Will Tong, a former equity investor at Citadel, and Stefan Raghavan, a former engineer at SpaceX.

High-Level Overview

Formula Insight is an AI-powered investment research platform designed to help institutional investors manage, track, and query their complex Excel-based financial models at scale. The company builds tools that unify financial model analysis with external research documents—such as SEC filings, earnings transcripts, and consensus estimates—into a single, searchable, and secure environment. By enabling users to search millions of Excel cells using plain English and cross-reference key performance indicators (KPIs) across hundreds of documents, Formula Insight dramatically accelerates the equity research workflow and strengthens the fact patterns behind investment decisions.

The platform primarily serves hedge funds, long-only asset managers, and other institutional investors who rely heavily on Excel for modeling but struggle with fragmented workflows between spreadsheets and research materials. It solves the critical pain point of lost insights when connecting financial models to supporting documents, reducing manual effort and improving decision quality. With backing from Y Combinator and integration with FactSet data, Formula Insight is gaining traction as a next-generation research infrastructure layer for public market investors, positioning itself as an essential tool in a competitive, data-intensive landscape.

---

Origin Story

Formula Insight was founded in 2024 as part of Y Combinator’s Summer 2024 cohort by Will Tong, a former equity investor at Citadel, and Stefan Raghavan, an aerospace engineer with experience at SpaceX. The idea emerged from Tong’s firsthand experience managing large portfolios using intricate Excel models, where he repeatedly encountered inefficiencies in linking model outputs to source documents like 10-Ks, earnings calls, and research reports. Despite the sophistication of these models, critical insights were often buried or disconnected from the broader research context.

Raghavan brought deep technical expertise in systems engineering and data infrastructure, allowing the duo to architect a platform that could parse and index vast Excel workbooks while preserving their structure and logic. Their combined background—one rooted in high-stakes investing, the other in scalable engineering—enabled them to build a solution that speaks both to the needs of portfolio managers and the realities of complex financial data. Early traction came quickly through direct engagement with hedge funds frustrated by clunky, siloed research workflows, validating the demand for a unified, queryable layer over institutional Excel data.

---

Core Differentiators

Unified Research & Modeling Layer
- Seamlessly connects internal Excel financial models with external documents (filings, transcripts, estimates), enabling cross-document inference and richer fact patterns.
- All findings are linked back to source documents, ensuring traceability, auditability, and accuracy.

Natural Language Search Over Excel
- Allows users to search millions of Excel cells using plain English queries (e.g., “Show me revenue growth assumptions for 2025 across all tech models”).
- Dramatically lowers the learning curve compared to traditional model navigation or SQL-based tools.

Granular KPI Analysis & Benchmarking
- Enables tracking of forecast changes over time, measurement of projection accuracy vs. actuals, and custom benchmarking across models and sectors.
- Supports plotting financial projections against consensus estimates to surface outliers and alpha-generating insights.

Built for Institutional Security & Scale
- Designed specifically for hedge funds and asset managers, with a secure environment that keeps sensitive financial models and forecasts protected.
- Integrates with FactSet for real-time market data and consensus estimates, enhancing analytical depth.

Workflow Efficiency & Speed
- Reduces hours of manual model review and document sifting into minutes of targeted analysis.
- Centralizes a firm’s most valuable Excel data into a single, queryable repository, preventing knowledge loss across teams and tenures.

---

Role in the Broader Tech Landscape

Formula Insight sits at the intersection of three powerful trends: the enduring dominance of Excel in finance, the rise of AI-augmented knowledge work, and the growing need for structured, searchable data layers within investment firms. Despite advances in SaaS and analytics platforms, Excel remains the de facto standard for financial modeling in institutional investing—especially in hedge funds—due to its flexibility, control, and deep integration with financial databases. However, this reliance creates a “dark data” problem: vast amounts of valuable assumptions, forecasts, and KPIs live in unindexed, siloed spreadsheets.

By applying modern search, AI, and document understanding techniques to this entrenched workflow, Formula Insight is helping modernize the research stack without forcing a shift away from Excel. The timing is critical: as markets become more competitive and information cycles accelerate, firms can no longer afford inefficient, manual research processes. Regulatory scrutiny, internal governance, and the need for reproducible decision-making further amplify demand for traceable, auditable research workflows.

In the broader ecosystem, Formula Insight represents a new class of “infrastructure for alpha”—tools that don’t replace analysts but augment their ability to find, validate, and act on insights faster. It also reflects a broader trend of engineers and operators from outside traditional finance (e.g., aerospace, tech) applying systems thinking to long-standing inefficiencies in financial services.

---

Quick Take & Future Outlook

Formula Insight is well-positioned to become the default research layer for institutional equity investors who refuse to abandon Excel but want the benefits of a modern, AI-native research stack. In the near term, expect deeper integrations with other data providers (Bloomberg, CapIQ), expanded support for alternative data sets, and enhanced collaboration features for research teams. Over time, the platform could evolve into a central “memory” for investment firms—preserving institutional knowledge, tracking analyst accuracy, and even powering automated idea generation and risk alerts.

As AI continues to reshape knowledge work, tools like Formula Insight will increasingly define competitive advantage in asset management: not by replacing human judgment, but by making it faster, more rigorous, and more scalable. The real prize isn’t just better search over spreadsheets—it’s turning a firm’s collective modeling work into a structured, reusable asset. For investors, that means fewer missed insights, stronger theses, and a clearer edge in the public markets.