
Tick Stock Panel is an open-source, self-hosted quantitative trading workstation for China's A-share market that combines high-performance backtesting with LLM-assisted strategy customization and stock analysis.
Tick Stock Panel (TSP) is an open-source, self-hosted quantitative workbench designed for retail traders and researchers operating in the Chinese A-share market. Powered by a modern stack featuring FastAPI, Polars, DuckDB, vectorbt, and React, TSP integrates stock screening, real-time market monitoring, and vectorized backtesting into a unified dashboard. The platform incorporates large language models to streamline natural-language strategy customization, post-market review generation, and multi-factor stock analysis, while providing an extensible plugin architecture for third-party market data providers.
TSP strikes a compelling balance between rigorous data engineering and modern AI copilot ergonomics, positioning itself as a standout self-hosted toolkit for retail quants.
* Modern data engine: Leveraging DuckDB, Polars, and vectorbt sidesteps traditional Pandas performance bottlenecks for vector backtesting and screening.
* Grounded LLM assistance: Instead of making speculative alpha claims, the AI integrations focus on developer ergonomics, strategy translation, and qualitative market review.
* Pluggable data architecture: Decoupled data provider support enables users to swap between standard feeds and custom sources like Baostock or TickFlow.
* Privacy-first self-hosting: Running entirely local or via Docker protects proprietary trading logic and avoids reliance on subscription-based cloud SaaS platforms.
DISCOVERED
1h ago
2026-09-25
PUBLISHED
1h ago
2026-09-25
RELEVANCE