AI algo trading - ai virtual stocks portfolios | LLM managed stocks portfolios

Check existing website: www.odin500.com Odin500 is a platform where AI models (ChatGPT, Claude, Gemini and Grok) manage virtual stock portfolios on the S&P 500, Dow Jones and Nasdaq-100. We track portfolios performance and allow users to copy and follow profitable portfolios. Odin500 includes Odin AI portfolios, as well as user-generated portfolios (UG). Users can pick a direction (long, short or long-short), criteria (technical, fundamental or news-momentum) and a rebalance schedule (daily, weekly or monthly). The AI chooses the stocks, and we track the results publicly. We have tens of portfolios live created by Claude, Gemini, ChatGPT, and Grok. We are looking for more succesfull portfolios. We need an AI expert to create investment strategies that performs better and more consistently. Design investment strategies framework: stock selection rules, position sizing, risk controls (stop-losses, drawdowns), rebalance cadence, and when to be long, short or in cash. Design prompt AI models and what market data we give them, so decisions are consistent and explainable. Define a simple baseline strategy and a testing method (A/B cohorts, benchmark comparison) so we can tell whether the AI adds value. Write a clear implementation spec for our developers can build from. Deliverables: A written diagnosis of the current performance with the top causes ranked. A strategy and risk-rules document. Revised prompt templates for each AI model. A testing and evaluation plan with metrics (return vs benchmark, drawdown, win rate, turnover). One or two review calls after our developers implement it. Required skills Hands-on experience with quantitative or systematic equity strategies, such as factor investing, momentum or mean reversion. Practical LLM prompt engineering for structured outputs, with working knowledge of OpenAI, Anthropic, Google and xAI models. Backtesting and statistics: avoiding overfitting, look-ahead bias and survivorship bias. Portfolio construction and risk management. Python and SQL for reviewing our data. Clear written communication, because you'll write specs for non-quants. Nice to have: a background in AI-driven investing products or research, and experience with fintech compliance (this is education and simulation, not investment advice). Please include in your proposal Relevant strategies or AI-in-finance projects you've worked on, with links if public. How you would first diagnose a set of portfolios that are all slightly negative. Your approach to telling skill from luck over a short track record. Your timeline and fixed price, or hourly rate.