Show HN：QuantDinger – 一個開源、本地化的 AI 量化交易平台

QuantDinger is an open-source, local-first AI-powered quantitative trading platform that I’ve been building for about six months. It’s designed to cover the full quant workflow — from research and strategy development to backtesting and live execution — while keeping everything running locally.

Most existing quant tools are cloud-based, which means strategies, indicators, and API keys often need to be uploaded to third-party servers. QuantDinger takes a different approach: it is local-first by default, so strategy logic and credentials stay on your own machine.

The platform currently supports multiple markets, including US equities, A-shares, Hong Kong stocks, crypto, forex, and futures.

Key features:

• Local-first architecture with Docker-based deployment
• AI-assisted strategy and indicator generation
• Python-native strategy development
• Visual indicators and K-line (candlestick) execution
• Backtesting and live trading support
• Multi-user support for self-hosted setups

QuantDinger is fully open source under the Apache 2.0 license and can be used commercially.

Demo:
https://ai.quantdinger.com

GitHub:
https://github.com/brokermr810/QuantDinger

I’d really appreciate feedback from people who’ve built or used trading systems, especially around architecture, backtesting design, and practical usability.

In my own trading and research, I found it hard to trust hosted quant platforms with proprietary strategies and exchange API keys. QuantDinger was built so that execution logic, credentials, and most data processing never leave the local environment.

The AI components are used mainly for research assistance (e.g. generating strategy ideas or indicators), not for opaque “black-box” execution. All strategies remain inspectable Python code.

Happy to answer technical questions about how the backtesting engine works, how markets are abstracted, or how the AI agents are integrated.

reply