Phoenix

Full-stack engineer, AI builder, data infra.

I build practical AI products and data systems from idea to production: product interfaces, backend services, automation, and the pipelines that keep them useful.

Beijing, China Building independent AI products
View work Follow on X

Products in the open.

Current and past projects sit where AI, public data, automation, and product design meet.

Polysage interface preview

Polysage

A public intelligence surface for Polymarket odds: opportunity markets, large-trade flow, movers, and model watchlists designed for daily scanning.

prediction markets data product market intelligence
AI X product preview

AI X

An AI-assisted content tool for generating, selecting, and scheduling Twitter/X posts with a practical workflow around creator-owned API keys.

AI content automation creator tooling

Builder first.

I work across the full stack because useful AI products rarely stop at one layer. The interface, backend, data model, observability, and delivery path all have to line up before a product becomes something people can return to.

01
Ship the working loop

Start with the smallest real workflow that creates signal, then harden it with feedback from production data.

02
Make data usable

Data infra is only valuable when people can trust it, inspect it, and turn it into decisions without ceremony.

03
Keep products practical

The best AI features disappear into the workflow instead of asking users to babysit a demo.

A small personal note.

Outside work, there is Nineteen, a golden British Shorthair born on August 19, 2022. He keeps the house warm, noisy in the right way, and a little less optimized.