deeter-analytics - Head of Algo Trading (AI)

• We hire for demonstrated ability and how you think, not for pedigree. You have a track record of building and shipping AI trading systems end to end, and strong quantitative and technical depth, but whether that came from a STEM degree, applied or research work, competitions, or a self-taught path does not matter to us. We welcome strong non-traditional backgrounds, and look for evidence that you are: • A strong first-principles thinker with real technical depth and sound judgment about what matters. • Rigorous about results, candid about uncertainty, and quick to revise your view in light of evidence. • Low ego and resilient: open to challenge, and comfortable being wrong in pursuit of the right answer. • Able to take a broad mandate, define the work that needs doing, and deliver. • Curious and genuinely energized by the problem. • AI-native. Fluent in modern AI, and able to use it to accelerate research, generate and test ideas, and support decisions across the team. You need not be the firm's deepest AI researcher, but you should be genuinely fluent and able to direct that work. • Empirical and iterative. You establish baselines, run efficient experiments, and compound learnings over time, while holding live-capital work to a high standard, guarding against overfitting and spurious signals. • Pragmatic. You select the right method for each problem and prioritize what moves P&L, rather than complexity for its own sake. • Clear and collaborative. You communicate concisely, give and receive feedback well, and work effectively across teams. • Modeling, classical and frontier. Deep facility with statistical and machine-learning modeling (forecasting, signal generation, risk) and with getting real leverage out of frontier AI models, plus the judgment to know which fits a problem and where newer methods (deep learning, reinforcement learning) add edge versus just overfit. • Experiment design & data sense. You design fast, cheap experiments, establish baselines, and will do things that don't scale to find early signal, turning messy and alternative data (e.g., social or text) into something usable, and reading it directly rather than hiding behind a single statistic. • Signal-versus-noise judgment. You can separate real edge from a fragile result or a false positive (honest out-of-sample testing plus an instinct for what's overfit) before anything touches live capital. • Operational build-out. You can get an MFT book live end to end (data, execution, deployment), making sensible buy-vs-build calls. • Prototype to production. A strong, hands-on coder (Python and the modern data/ML stack) who takes an idea from prototype to production quickly, using AI to move faster.