As AI labs develop and evaluate frontier models, they need real-world production codebases and commit histories written by human software engineers. Through Andela’s partnership with Pangea.ai, you can license qualifying private repositories to leading AI research labs, earn recurring non-exclusive payouts, and retain 100% ownership of your intellectual property.
1. Why License Your Code?
- Retain full ownership: Licenses are 100% non-exclusive. You keep your copyright and IP, and you can continue using, modifying, and deploying your code freely.
- Earn recurring revenue: Because licenses are non-exclusive, the same codebase can be licensed to multiple AI labs over time as buyer demand arises.
- Zero cost, zero lift: Joining is completely free. No fees are deducted from your payout share, and you do not need to rewrite or clean up your code.
- Direct contribution to AI progress: Production-grade software with genuine git histories (showing bug fixes, refactoring, and feature additions) is essential training data that cannot be generated synthetically.
2. Crucial Ownership Rules: What You Can & Cannot Submit
⚠️Important Legal Notice: You may only submit codebases that you own outright or have explicit written legal authority to license.
Eligible (Submit These) |
Ineligible (Do NOT Submit) |
| Personal side projects, SaaS platforms, or internal tools you built | Code written for Andela clients |
| Code from your own startups or personal R&D | Code written for an employer under an employment agreement |
| Freelance or contract client work where you retained all IP rights | Client work where IP was transferred or assigned to the client |
3. What Qualifies?
AI labs have specific criteria for the repositories they accept into evaluation and training pipelines:
✅ Strongest Fit
- Strictly private: Code that is not currently on public GitHub/GitLab/Bitbucket, has never been public in the past (including Internet Archive snapshots), and was never distributed on package registries (such as npm, PyPI, or Maven).
- Human-written: Code created primarily before 2025 represents the strongest fit. (Code written from 2025 onwards remains eligible but is subject to review for AI generation).
- Substantial applications: Full backends, APIs, SaaS tools, and multi-service systems with meaningful complexity.
- Rich Git history: Multi-developer or multi-year commit histories, pull requests, issue threads, and automated tests directly connected from your Git host. Flattened repositories lose most of their licensing value.
❌ Ineligible / Declined
- Public repositories or open-source forks/mirrors.
- AI-generated code (labs enforce strict zero-tolerance policies).
- WordPress sites, themes, and generic plugins.
- Raw datasets, data dumps, or repositories composed almost entirely of vendored third-party dependencies and build artifacts.
4. How It Works: Step-by-Step
[1. Get Free Estimate] ➔ [2. Sign Non-Exclusive Agreement] ➔ [3. Automated Scrubbing] ➔ [4. Licensing & Payout]1. Get an instant, risk-free valuation
Sign up via here and connect GitHub, GitLab, or Bitbucket with read-only access. You’ll receive an automated estimate in minutes.
(Prefer not to connect online? You can run the open estimation script locally on your machine—no code ever leaves your environment.)
2. Review and sign once
When you are satisfied with the estimates, sign the unified licensing agreement. One signature covers all current and future repository licensing. You choose exactly which repositories to include.
3. Automated security scrubbing
You do not need to spend time sanitizing your codebase. Pangea clones a separate copy and automatically scrubs API keys, environment variables, passwords, personal identifiable information (PII), and contributor identities before labs receive access.
4. Licensing & payment
Repos are batched and matched against buyer demands. When a license transaction closes and the buyer remits payment, you receive your funds within 15 days.
⚠️Any agreements signed are between you and Pangea.ai. Andela is not liable or responsible for any conditions in the agreement.
5. FAQ
- Do I lose intellectual property rights? No. The license is strictly non-exclusive, meaning you remain the sole owner of the codebase and may continue building, deploying, selling, or commercializing it.
- Do I need to clean up my git history or remove secrets first? No. Pangea handles secrets stripping and PII scrubbing on a duplicate copy. Altering or squashing your git commit history manually can actually diminish the valuation of your codebase.
- What if I change my mind later? You can withdraw any codebase at any point before it is presented to buyers. Once licensed, you can opt out of any future licensing rounds by providing 30 days’ notice.
- Is my participation public? No. Contributor information is pseudonymized during the grading process, and your involvement is never publicly shared or marketed without your express written consent.
-
Can I upload tickets, task boards, or documentation? Yes. Pangea currently supports an experimental (beta) upload feature for supplementary context—including Jira/Linear tickets, Notion/Confluence docs, and Slack exports. This data is sanitized, securely reviewed, and deleted after ingestion.
Note: Operational data uploads are currently experimental and do not offer guaranteed payouts.
Ready to Get Started?
Check your codebase eligibility and view your valuation in minutes: 👉 Connect Your Repos
For additional questions or support, please reach out to the Head of Talent Experience at Pangea Jazlin Rosewitz