Repolode
Submit
lc2panda

lc2panda/alphastream

An AI-driven investment analysis platform where multiple LangGraph agents analyze A-shares, US stocks and ETFs, streaming each step of the decision process over SSE.

View on GitHub
Stars
898
Forks
202
Open issues
5
Last commit
3 months ago

Repolode describes open-source projects and does not give investment advice. Trading and investing can lose money: assess the risk yourself and follow the laws where you live before using this project.

What is it?

alphastream is a full-stack AI investment analysis platform built with Next.js 16, React 19, Flask 3.1 and LangGraph. It coordinates 14 specialized agents covering technical analysis, fundamentals, capital flows, sentiment, bull and bear debate, risk and decision-making, plus four investor persona agents. Agents call data tools through OpenAI-compatible function calling, so any compatible LLM provider can be used, with DeepSeek as the documented default.

The data layer pulls from AKShare and BaoStock for A-shares with failover between them, and also connects to SEC EDGAR, yfinance, FRED, ccxt and several RSS feeds. A web search layer aggregates 17 engines, and an OpenCLI-based scraper collects discussions from Xueqiu, Eastmoney forums and Cailian Press. Users see charts, watchlists, portfolios and ten artifact cards such as K-line, risk radar and decision cards.

The system streams agent events (start, progress, thinking, tool calls, completion) over Server-Sent Events, so the front end can show the reasoning as it happens. It deploys with Docker Compose behind Nginx, with Redis as an optional cache. The project states that its output is not investment advice.

Key features

  • 01Coordinates 14 LangGraph agents with parallel, sequential and conditional routing.
  • 02Streams agent progress, reasoning and tool calls to the browser over SSE.
  • 03Covers A-shares, US stocks, Hong Kong stocks, ETFs and indices.
  • 04Provides K-line, moving average, MACD, RSI and Bollinger Band indicators.
  • 05Aggregates 17 search engines and multiple news sources with sentiment extraction.
  • 06Works with any OpenAI SDK-compatible LLM endpoint.

Use cases

  • Exploring how an LLM multi-agent workflow can be applied to market data analysis.
  • Studying a reference implementation of SSE-streamed agent reasoning in a Flask and Next.js stack.
  • Prototyping a self-hosted stock research dashboard for A-shares and US equities.
  • Experimenting with LLM function calling over financial data adapters.

Quick start

  1. 1

    Clone the repository

    git clone https://github.com/lc2panda/alphastream.git
  2. 2

    Enter the project directory and copy the environment template

    cd alphastream && cp .env-example .env
  3. 3

    Fill in OPENAI_API_KEY and related model settings in .env

  4. 4

    Start all services with Docker Compose

    docker compose up -d

Things to know

  • Requires an LLM API key (DeepSeek or another OpenAI-compatible provider) to run the AI agents.
  • Requires a Docker or Node.js and Python setup, and multiple services such as Redis and Nginx, so self-hosting takes some effort.

In collections

Open-source AI agent frameworks & apps

Topics

#ai#llm#langgraph#multi-agent#stock-analysis#fintech#sse#flask

FAQ

What is alphastream?
An AI-driven investment analysis platform where multiple LangGraph agents analyze A-shares, US stocks and ETFs, streaming each step of the decision process over SSE. alphastream is a full-stack AI investment analysis platform built with Next.js 16, React 19, Flask 3.1 and LangGraph. It coordinates 14 specialized agents covering technical analysis, fundamentals, capital flows, sentiment, bull and bear debate, risk and decision-making, plus four investor persona agents. Agents call data tools through OpenAI-compatible function calling, so any compatible LLM provider can be used, with DeepSeek as the documented default.
Is alphastream free to use? What license does it have?
alphastream is open source under the MIT license. Check the license terms before using it in commercial products.
Is alphastream still maintained?
Updated within 6 months. The last commit to alphastream was on Aug 7, 2026.
What can I use alphastream for?
Exploring how an LLM multi-agent workflow can be applied to market data analysis. Studying a reference implementation of SSE-streamed agent reasoning in a Flask and Next.js stack. Prototyping a self-hosted stock research dashboard for A-shares and US equities. Experimenting with LLM function calling over financial data adapters.
How do I install alphastream?
Clone the repository: git clone https://github.com/lc2panda/alphastream.git Enter the project directory and copy the environment template: cd alphastream && cp .env-example .env Fill in OPENAI_API_KEY and related model settings in .env Start all services with Docker Compose: docker compose up -d
Where can I get alphastream?
The source code and installation instructions are on GitHub: https://github.com/lc2panda/alphastream

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects