Skip to content

CONTRIBUTING.md

Latest commit

 

History

History
121 lines (86 loc) · 2.47 KB

CONTRIBUTING.md

File metadata and controls

121 lines (86 loc) · 2.47 KB

Contributing to DocuQuery

Thank you for your interest in contributing to DocuQuery. This project is designed as a production-oriented RAG system, and we welcome contributions that improve its performance, reliability, and usability.

How to Contribute

1. Fork the Repository

Create your own fork of the repository and clone it locally.

2. Create a Feature Branch

Use a clear and descriptive branch name:

feature/
fix/
improvement/

Examples:

- feature/add-hybrid-search
- fix/retrieval-threshold-bug
- improvement/api-response-format

3. Make Changes

  • Keep changes focused and minimal
  • Follow existing project structure and naming conventions
  • Add comments where necessary for clarity

4. Write Tests

All contributions should include appropriate test coverage:

  • Unit tests for core logic
  • Integration tests for pipelines or APIs
  • Ensure all existing tests continue to pass

5. Run Tests

Validation Checklist:

Before submitting your changes, please ensure the following checks pass:

pytest
python -m app.run_pipeline
python -m eval.run_eval
uvicorn app.main:app --reload

6. Submit a Pull Request

When opening a PR, include:

  • Clear description of the change
  • Motivation for the change
  • Any relevant screenshots or logs (if applicable)

Pull Request Guidelines

  • Keep PRs small and focused
  • Avoid unrelated changes in a single PR
  • Ensure code is clean and readable
  • Respond to review comments promptly

All pull requests will be reviewed before merging.

Code Style

  • Follow consistent Python formatting
  • Use meaningful variable and function names
  • Keep functions modular and reusable
  • Avoid unnecessary complexity

Areas for Contribution

We actively welcome improvements in:

  • Retrieval quality (ranking, filtering)
  • Evaluation metrics and datasets
  • Performance optimizations
  • API improvements
  • Documentation clarity
  • Testing coverage

Reporting Issues

If you encounter a bug or have a feature request:

  • Use the GitHub Issues tab
  • Provide clear steps to reproduce (for bugs)
  • Include logs or error messages if available

Contribution Philosophy

DocuQuery is built with a focus on:

  • Reliability over shortcuts
  • Clarity over cleverness
  • Measurable improvements over assumptions

Contributions aligning with these principles are highly encouraged.

Author

Aditya Bisoyi
Software Engineer focused on scalable systems and applied machine learning