Claude-Powered Stock Trading Bot
Budget: $30 – $250 USD
I want a fully-functional trading bot built around Anthropic’s Claude models and tuned specifically for stock trading. The system must be capable of running day-trading, swing-trading and scalping strategies in parallel, letting Claude handle the pattern recognition, news sentiment and order timing while your code manages data feeds, risk controls and execution.
You’re free to recommend the most suitable brokerage or exchange APIs—NASDAQ, NYSE and even international venues are all on the table as long as the final solution can be extended easily. The important part is clean, well-documented code (Python preferred), a modular strategy layer that lets me switch models or tweak parameters, and a dashboard or simple CLI where I can monitor performance in real time.
Deliverables
• Source code and environment files
• Claude prompt/chain configurations for each strategy
• Back-test reports proving day, swing and scalping logic over at least one year of historical data
• Deployment guide so I can spin the bot up on my own VPS or cloud instance
Acceptance Criteria
1. Bot places live or paper trades end-to-end without manual intervention.
2. Each strategy runs independently and logs PnL, drawdown and trade rationale.
3. All external keys, endpoints and model parameters are configurable from a single file.
If you have prior experience marrying LLMs with algorithmic trading systems, that’s a big plus. Let me know how you’d structure the project and any libraries you consider essential.
You’re free to recommend the most suitable brokerage or exchange APIs—NASDAQ, NYSE and even international venues are all on the table as long as the final solution can be extended easily. The important part is clean, well-documented code (Python preferred), a modular strategy layer that lets me switch models or tweak parameters, and a dashboard or simple CLI where I can monitor performance in real time.
Deliverables
• Source code and environment files
• Claude prompt/chain configurations for each strategy
• Back-test reports proving day, swing and scalping logic over at least one year of historical data
• Deployment guide so I can spin the bot up on my own VPS or cloud instance
Acceptance Criteria
1. Bot places live or paper trades end-to-end without manual intervention.
2. Each strategy runs independently and logs PnL, drawdown and trade rationale.
3. All external keys, endpoints and model parameters are configurable from a single file.
If you have prior experience marrying LLMs with algorithmic trading systems, that’s a big plus. Let me know how you’d structure the project and any libraries you consider essential.