Statistician by training, applied scientist by trade. These days I build agentic AI systems that make working with data easier — and I'm lucky that my day job (data science at Paramount) is exactly what I'd be tinkering with anyway.

Three layers I care about:

• 🧠 Shared context — team knowledge that humans and agents can both operate on: capture it once, keep it true, let everyone (and every agent) build on it.
• 🔁 The dev loop — pain-free multi-agent development: coordination, review gates, and validation so a fleet of agents — and the humans steering them — ship together.
• 📈 The application layer — wiring LLM intelligence to domain expertise over real data: semantic contracts, deterministic pipelines with agents at the edges, and evaluation you can trust.

🌱 Roots: PhD in biostatistics. The older repos here — statistical methods in R / C++ / Python — are from that life. No longer maintained, but the statistical thinking never left.

💬 Open to collaborating on ML / AI / statistics projects · 📫 qianl90(at)gmail(dot)com