Launch Playbook

July 4, 2026 ยท View on GitHub

This playbook turns repository improvements into a public launch that is useful rather than noisy.

Pre-Launch Checklist

  • CI is green on main.
  • README quick start is still accurate.
  • The social preview image is uploaded in GitHub repository settings.
  • Repository topics are set.
  • The first release notes are ready.
  • At least one benchmark command is easy to copy.
  • The public launch checklist issue is open.

Description:

End-to-end PyTorch research stack for ML-enhanced multi-factor trading, factor tensors, bias correction, portfolio optimization, and reproducible backtesting.

Website:

https://arxiv.org/abs/2507.07107

Topics:

quantitative-finance, algorithmic-trading, machine-learning, pytorch, portfolio-optimization, alpha-factors, backtesting, research, quant, factor-models, financial-machine-learning

Social preview image:

docs/assets/social-preview.png

GitHub currently requires uploading the social preview image through the repository Settings page.

Launch Post Order

  1. Create a GitHub release.
  2. Post a short technical launch note on GitHub issue/discussion.
  3. Share the short social post.
  4. Share the longer LinkedIn/community post.
  5. Reply to questions and turn feedback into issues.

Launch Message

Use a concrete, reproducible promise:

Clone the repo and run a full synthetic factor-to-backtest pipeline with one command.

Avoid performance hype. Lead with reproducibility, engineering depth, and extensibility.

First Week Goals

  • 3 benchmark reports from different machines.
  • 1 public-data reproduction report.
  • 1 external issue that identifies unclear documentation.
  • 1 external PR, even if it is a small docs fix.

Where to Share Carefully

Share only where the audience is relevant and the post includes technical value:

  • quantitative finance communities
  • ML engineering communities
  • research software communities
  • university or student quant groups
  • personal LinkedIn/X account

Do not post the same text everywhere. Adjust the ask to the community:

  • For quant users: ask for factor/backtest feedback.
  • For ML users: ask for tensor benchmark results.
  • For students: ask for reproducibility feedback.
  • For engineers: ask for API, tests, and packaging feedback.