Skip to main content
Bridge Town

About Bridge Town

Financial planning that
speaks your language

Bridge Town replaces proprietary tools like Anaplan and Pigment with a bring-your-own-agent, code-first approach. FP&A analysts describe business logic in plain English — their agent handles the Python through Bridge Town MCP tools, while Bridge Town manages version control and collaboration.

/ What we're building

A platform where financial models are code — readable, testable, and version-controlled.

  • 01

    Agent-native modelling

    Describe your budget, forecast, or variance analysis in natural language. Your agent writes the Python through Bridge Town MCP tools; Bridge Town connects the data, runs the model, and versions the result.

  • 02

    Version control built-in

    Every model has a complete, versioned project history. Branch, diff, merge, and roll back — the same discipline that made software engineering reliable, applied to financial models.

  • 03

    Real collaboration

    Review model changes like code changes. Impact diffs show exactly how a proposed change affects your forecast before you merge.

  • 04

    Google Sheets integration

    Connect your existing Sheets as data sources. Drivers, actuals, and assumptions flow directly into your models without manual exports.

  • 05

    Interactive dashboards

    Models produce Plotly dashboards automatically. Share a link — stakeholders get a live view without touching a spreadsheet.

  • 06

    MCP-first interface

    Bridge Town is built around the Model Context Protocol. Use it via Claude Desktop, Claude Code, or any MCP-compatible AI client.

/ Our principles

The decisions behind how we build.

  • 1

    Open standards over proprietary lock-in

    Your models are plain Python. Your data pipeline runs on open tools. You own your work — export it, run it locally, or move on without a migration fee.

  • 2

    Reproducibility is non-negotiable

    Re-running a model at a past commit against the same snapshot produces the same output. That is the foundation of trustworthy financial reporting.

  • 3

    Your agent is the interface, not the gimmick

    Your own Claude, Codex, or Cursor agent translates analyst intent into well-structured, testable Python models through Bridge Town MCP tools. The agent is useful because the underlying platform is solid — not the other way around.

/ Why we built this

Enterprise FP&A tools are expensive, slow, and proprietary.

  • Reason 01

    Anaplan costs too much

    Enterprise FP&A tools charge six figures for features that add complexity without adding value. Bridge Town does more for a fraction of the cost.

  • Reason 02

    Spreadsheets do not scale

    Version control, peer review, and automated testing exist in every other field of knowledge work. FP&A deserves the same infrastructure.

  • Reason 03

    Analysts should not need to learn proprietary languages

    Pigment and Anaplan have their own formula languages that go nowhere. We use Python — the most widely-used language in data work — so analysts build transferable skills.

Contact

Get in touch