Automated multi-agent trading analysis with GPT-5, Telegram, Coinbase & Notion
π Overview This modern n8n workflow implements a multi-agent trading engine that orchestrates valuation, sentiment, fundamentals, technicals, and macro analyses to generate a single portfolio action with built-in ris...
Template notes
π Overview This modern n8n workflow implements a multi-agent trading engine that orchestrates valuation, sentiment, fundamentals, technicals, and macro analyses to generate a single portfolio action with built-in risk controls. It integrates Telegram for live commands, fetches market data, fans out to expert LLM agents, applies position limits via a Risk Manager, executes orders, logs to Notion, and sends a summary back to Telegram.
π§ What It Does - Telegram Trigger listens for ticker commands and context inputs π² - Market Data Node fetches live prices and exchange rates from an API (no hardcoded keys) π - LLM Agents run parallel analyses: valuation, sentiment, macro, fundamentals, technicals π - Risk Manager enforces max position sizes, stop-loss limits, and confidence thresholds βοΈ - Portfolio Manager aggregates signals into a final BUY/SELL/HOLD decision with allocation % π - Execute Order sends trade requests via HTTP Request π - Log to Notion writes a full audit trail for compliance π - Telegram Output posts a concise report with signals, risk notes, and final decision π€
π‘ Why Itβs Useful This template illustrates a modular βinvestment committeeβ architecture that is easy to extend, swap agents, and maintain. It follows n8nβs best practices for template submissions: sticky-note documentation, placeholder credentials, markdown descriptions, and clear H2 headings.
π Prerequisites - Telegram Bot credentials configured in n8n - Exchange or data API credentials stored as n8n Credentials (no inline keys) - OpenAI (or other LLM) API credential - Notion integration credentials
π οΈ How to Use 1. Import the JSON into n8n and open the canvas. 2. Read each Sticky Note for node-by-node setup tips and rate-limit guidance ποΈ 3. Configure credentials via the n8n Credentials Manager π 4. Test each branch (data fetch, agents, risk logic) in isolation before enabling order execution β
π Architecture Layers - Trigger: Telegram Trigger β - Data: HTTP Request β - Analysis: Parallel LLM Agents β - Risk: Risk Manager β - Decision: Portfolio Manager β - Action: Execute Order, Log to Notion, Send Telegram summary
π Security & Maintenance All API keys are stored securely as credentials. Sticky Notes document required scopes, retry strategies, and error-handling paths to ensure observability and safe testing.
Enjoy building and customizing your own AI-powered hedge-fund workflow!