Self-Hosted A-Share Quant Workbench: Stock Screener, Backtester, and Monitor in One Docker Container

Quantitative trading has always felt like an institutional game — expensive data feeds, complex infrastructure, and constant maintenance. The author of this project is clearly a retail trader too, because "zero-ops" is right there in the first line of the description.

What It Is

An A-share quant workbench built on TickFlow data. Three core functions: stock screener, backtester, and monitor. The tech stack is React + FastAPI + DuckDB + Polars — solid, modern choices for this type of application.

Deployment is refreshingly simple: one Docker command and you're running. There's even a Free mode that works without registration. If you want extended data, just add a TickFlow API key — and the author provides an invite code (`V3KDKGXPEA`) for free access to sector/industry datasets.

Technical Highlights

• **DuckDB embedded analytics**: No separate database to set up. Data lives in-process, queries are fast.
• **Polars for data processing**: Significantly faster than Pandas for large candlestick datasets.
• **Capability-driven tiers**: Features automatically adapt to your TickFlow subscription level. Free users aren't bombarded with paywall messages.
• **Third-party data integration**: Supports Tushare and other data sources. Configurable via the UI without code changes.

Why It Matters

The most obvious commercial path here would be SaaS-ifying it — but the author chose open source, which is probably the faster route to traction. In the quant tools space, word-of-mouth depends heavily on data quality and transparency. TickFlow as a data source has decent street cred in the Chinese trading community.

For retail traders, the biggest value prop is approachability: the Dashboard surfaces key indicators out of the box, and the Screener supports natural-language strategy descriptions (with an AI key configured).

Currently TickFlow-only for data. Switching data sources requires modifying field mappings. The author says they'll build a data-source switcher if there's enough demand.

221 stars in 6 days tells me this hit a real, widespread pain point among Chinese retail traders.

docker run -d -p 3000:3000 --name stock-panel shy3130/tickflow-stock-panel