Fundamental Research Labs develops autonomous, collaborative, and socially intelligent digital human agents. Their initial product, Shortcut, applies agentic reasoning to spreadsheet intelligence, acting as an AI analyst for data modeling and simulation. The company focuses on building deeply capable systems to advance human-AI collaboration.
Funding
Funding not disclosed
Founders
Product
Problem
Organizations rely heavily on spreadsheets for data modeling, forecasting, and scenario analysis, but building accurate models and running simulations manually is time‑consuming, error‑prone, and often lacks transparent audit trails. Non‑technical users struggle to extract insights from raw data, while expert analysts spend disproportionate effort on repetitive spreadsheet tasks.
Solution
Shortcut provides an autonomous digital assistant that operates directly within Excel to perform agentic reasoning over spreadsheet data. Leveraging a multi‑agent architecture, the system interprets raw inputs, generates statistical or predictive models, and answers natural‑language queries about the data. It can execute parallel simulations and present results with full step‑by‑step traceability, enabling users to verify assumptions and reproduce outcomes. By handling routine analysis tasks, Shortcut frees analysts to focus on higher‑level decision making and reduces the risk of manual errors. The assistant integrates with existing spreadsheet workflows, requiring no separate platform or extensive training.
Target Audience
Primary users are business analysts, financial modelers, and product managers who rely on Excel for data‑driven decision making, as well as enterprise teams seeking to automate routine spreadsheet analytics.
Features
- AI‑driven Excel add‑in that parses raw tables and automatically constructs appropriate analytical models (e.g., regression, time‑series, Monte Carlo).
- Multi‑agent reasoning engine that decomposes complex queries into coordinated sub‑tasks for faster, parallel execution.
- Natural‑language interface allowing users to ask “What‑if” questions and receive concise, data‑backed answers within the spreadsheet.
- Full audit log that records each transformation, model selection, and simulation step for reproducibility and compliance.
- Real‑time collaboration support so multiple users can view and interact with the assistant’s outputs simultaneously.
- Secure cloud‑backed compute layer that scales processing power while keeping spreadsheet data encrypted in transit and at rest.
- API hooks for integrating the assistant’s capabilities into custom enterprise analytics pipelines.