Daily Stock Analysis — LLM-Powered Multi-Market Stock Intelligence System

Introduction

Daily Stock Analysis is an open-source Python system that uses large language models to perform automated stock analysis across A-stock, US, and Hong Kong markets. It aggregates multi-source data, processes real-time news, and generates comprehensive reports with actionable insights.

What It Does

• Fetches and consolidates stock data from multiple financial data sources
• Integrates with LLMs to generate natural-language analysis reports
• Monitors real-time news and market events affecting stock performance
• Produces decision dashboards for visualizing trends and signals
• Sends automated notifications when key signals trigger
• Supports A-stock, US, and Hong Kong markets in a unified workflow

Architecture Overview

Built in Python with a modular pipeline: data collectors fetch market data and news, a processing layer normalizes the data, LLM modules generate structured reports, a dashboard layer renders results, and a notification service pushes alerts. Configuration is managed through YAML files.

Self-Hosting & Configuration

• Requires Python 3.8+ with pip
• Configure API keys for data providers and LLM service in config.yaml
• Supports OpenAI-compatible APIs and local models
• Schedule runs via cron or system task schedulers
• All data stays on your infrastructure

Key Features

• Multi-market coverage: Chinese A-shares, US equities, Hong Kong stocks
• LLM-generated contextual interpretation beyond raw numbers
• Multi-source data aggregation reducing single-source bias
• Customizable alerts on price movements, volume spikes, or news
• Report archiving for historical comparison

Comparison with Similar Tools

• FinGPT — financial NLP research focus; this tool provides a ready-to-use pipeline with dashboards and notifications
• OpenBB Terminal — interactive terminal; this tool emphasizes automated daily reporting with LLM summaries
• Qlib — Microsoft's ML-driven trading platform; this tool targets daily intelligence reports

FAQ

Q: Which LLM providers are supported? A: OpenAI-compatible APIs, local models, and other providers with a compatible interface.

Q: Is it suitable for automated trading? A: No. It provides analysis and reports to inform decisions but does not connect to brokerage APIs.

Q: Can I add custom data sources? A: Yes. The data collector layer is modular with an adapter pattern for new sources.

Sources

• https://github.com/ZhuLinsen/daily_stock_analysis