QuantLabs demos AI trading bots
- QuantLabs founder Bryan Downing published a post and companion video saying he spent the past five to six weeks running “100% AI-generated” Python trading bots in CME futures markets. - Downing said his workflow scrapes news from more than 30 sources, turns it into a 25-page PDF, then uses prompts to generate 8 to 12 strategies a day. - QuantLabs also points users to public GitHub repos with Python trading code, but the materials provide limited independently verifiable performance detail. (quantlabsnet.com)
Trading bots are software that place trades automatically, and QuantLabs says artificial intelligence can now write those bots from plain-English prompts. Bryan Downing said this week he has been running “100% AI-generated” Python bots for five to six weeks. (quantlabsnet.com) (youtube.com) Downing said the system is focused on Chicago Mercantile Exchange futures, not individual stocks. He wrote that the workflow starts by collecting near real-time market news from more than 30 sources. (quantlabsnet.com) He said that news is compiled into a 25-page PDF on futures and options markets, and that the PDF is then used as context for code generation. From there, he said, the system can produce 8 to 12 trading strategies in a single day. (quantlabsnet.com) In plain terms, the pitch is that a language model acts like a junior programmer: it reads market context, writes Python, and revises the code after feedback. QuantLabs says users can do that inside Visual Studio Code with Anthropic’s Claude models and other tools. (quantlabsnet.com) (youtube.com) QuantLabs’ public materials frame that as a replacement for manual strategy coding. Downing wrote that he no longer hand-codes and is using English prompts to have the model write, analyze, and fix scripts. (quantlabsnet.com) The company’s YouTube channel pushes the claim further, describing “11 bots running real-time” and “700+ lines of production-grade trading bot code in minutes.” A separate video title claims a “100% win-rate gold strategy.” (youtube.com 1) (youtube.com 2) QuantLabs also has public GitHub repositories tied to the pitch. One repository, `ai-powered-trading`, shows Python files including `app.py` and `trading_engine.py`, and another repository lists four trading-strategy scripts. (github.com 1) (github.com 2) Those repositories show code artifacts, but the public pages surfaced here do not show a full audited track record, execution assumptions, fees, or slippage analysis. The post and videos describe profits and win rates, while the visible public evidence is mostly code, marketing copy, and high-level workflow descriptions. (quantlabsnet.com) (github.com) (youtube.com) That leaves the core claim in a familiar place for algorithmic trading: easy to demo, harder to verify. QuantLabs has shown how it wants AI to fit into bot building, but outside teams would still need reproducible tests before treating the output as investable. (quantlabsnet.com) (github.com)
