Skip to content

Repository files navigation

Cogence

Engineering Intelligence Platform

Transform engineering activity into clear, actionable, business-readable intelligence.

Overview

Cogence automatically analyzes signals from software development teams and converts them into concise reports that help non-technical leaders understand what engineering accomplished, what's happening, and what requires attention.

The Problem

Engineering teams create enormous amounts of information daily through commits, deployments, pull requests, and discussions. Yet most non-technical leaders cannot easily understand this information, creating a gap between those building software and those making business decisions.

The Solution

Cogence bridges that gap by continuously analyzing engineering signals and generating business-readable reports that answer:

  • What did engineering accomplish?
  • What are teams working on?
  • Where are risks emerging?
  • What should leadership know?

MVP v1 Pilot

The pilot focuses on one question: "What did we do during the last 24 hours?"

Data Sources: Gitea repositories and Git commits only

Delivery: REST API + Rocket.Chat at 21:00 Asia/Tehran (no dashboard)

Output: Daily business-readable report with:

  • Executive summary
  • Active repositories
  • Contributor activities
  • Management notes

See MVP v1 Documentation and Product Slices for build order.

Documentation

Product

Architecture

Development

Architecture Decision Records (ADRs)

  • ADR-001 - Commits as source of truth
  • ADR-002 - Business language in reports
  • ADR-003 - No code analysis in MVP
  • ADR-004 - Signals over surveillance
  • ADR-005 - AI generates summaries
  • ADR-006 - Daily report first
  • ADR-007 - Human readability priority
  • ADR-008 - Single-tenant approach
  • ADR-009 - FastAPI backend choice
  • ADR-010 - PostgreSQL as system of record
  • ADR-011 - Scheduled data collection

Technology Stack

  • Backend: FastAPI (Python)
  • Database: PostgreSQL
  • Data Source: Gitea API
  • AI: LLM for report generation
  • Delivery: REST API + Rocket.Chat
  • Deployment: Single-tenant, internal-first

Core Principles

  1. Business First - Reports for managers, not engineers
  2. Signals Over Surveillance - Understand work, don't monitor people
  3. Context Before Metrics - Explain what happened, not just numbers
  4. Explainability Over Mystery - Transparent AI summaries
  5. Human-Centered Reporting - Readable in under 60 seconds

Project Status

🚧 In Development — MVP v1 Pilot

Current focus: Product Slice 1 — collect commits from Gitea.

License

See LICENSE file for details.

Contributing

We welcome contributions! Please read CONTRIBUTING.md for guidelines.

Support

For questions or issues, please open a GitHub issue.

About

No description, website, or topics provided.

Resources

Contributing

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages