Open intelligence infrastructure for scientific discovery.

πAI Lab is the scientific-intelligence and open-technology organization initiated by the πHub big-science program. Its mandate is deliberately cross-disciplinary: it is not limited to proteomics, biomedicine, or any single scientific field.

We develop evidence-grounded, reproducible systems that help researchers move from literature and data to analysis, testable hypotheses, and validated results.

• Scientific agents and research workflows
• Evidence, literature, and knowledge infrastructure
• Multimodal scientific data and computing systems
• Evaluation, provenance, and reproducibility tooling
• Open interfaces, standards, and reusable research software

Our work may span life sciences and medicine, chemistry and drug discovery, materials and energy, physics and engineering, Earth and environmental sciences, and general scientific infrastructure.

• Evidence before confidence. Claims must remain traceable to data, sources, or explicit assumptions.
• Reproducibility by design. Environments, inputs, methods, and outputs should be inspectable and repeatable.
• Human accountability. AI can extend scientific work, but named people remain responsible for decisions and releases.
• Open interfaces. Components should be composable, portable, and useful beyond one project.
• Cross-disciplinary transfer. We look for abstractions that travel across scientific domains without erasing domain constraints.
• Maintenance is part of research. A published tool without ownership, documentation, or a lifecycle is unfinished work.

Not every experiment becomes an official πAI Lab project. Projects are admitted only when they have a clear scientific purpose, named maintainers, reproducible entry points, an explicit license, documented evidence boundaries, and a credible maintenance path.

See our governance, project policy, and contribution guide.

πAI Lab is in its founding stage. The initial portfolio is being curated; repositories will be added when they meet the organization's admission baseline rather than to make the organization appear artificially full.

We welcome researchers, engineers, maintainers, and scientific institutions who share these standards. Use the issue tracker of the relevant project for technical work, or contact liuzaoqu@163.com for organization-level collaboration.

Guangzhou, China

πAI Lab · 面向科学发现的开放智能基础设施