Python model files instead of proprietary planning formulas.
Built for FP&A workflows
What the model workspace gives your team
Bridge Town is the versioned product layer beneath your AI agent: model storage, collaboration, run outputs, and review history that agents and finance teams can both understand.
Project history that records model changes, runs, and output revisions.
Readable assumptions and source data that an MCP-compatible agent can inspect before making edits.
Team-facing review surfaces for logic, outputs, and commit history.
How it works
Model workflow
- 1
Connect your own agent to Bridge Town through MCP.
- 2
Ask the agent to create or revise model logic in plain English.
- 3
Review the generated Python, assumptions, and run outputs before adopting the change.
- 4
Keep the accepted model version in the project history for later comparison.
Common use cases
Where teams use this page of the product
Revenue forecastsHeadcount modelsCash runway modelsBoard plan packages
Frequently asked questions
Product details
Does Bridge Town run the language model for me?
No. Bridge Town is bring-your-own-agent by design. Your Claude, Codex, or other MCP-compatible agent connects to Bridge Town and performs the modeling work with the access you grant.
Are models stored as spreadsheets?
Bridge Town models are versioned project files, with Python model logic and captured outputs that can be reviewed like a codebase.
Connect your agent to Bridge Town
Start with a versioned workspace, then let your own MCP-compatible agent build, inspect, and revise the finance model with you.
Start Free