Head of Applied AI & Trading Systems (Iron Man Track)

8 months ago

Remote, WorldwideStaff+

Responsibilities

Design and maintain automated pipelines that ingest, clean, normalize, and alert on market and alternative data.
Build LLM-powered systems for summarizing, extracting, comparing, and reasoning over financial research and filings.
Create natural-language tools for querying proprietary research and fast discretionary back-testing.
Develop sentiment, relevance, and entity-recognition filters that map events to positions, watchlists, and risk exposure.
Build dashboards surfacing market regimes, anomalies, and dislocations rather than vanity metrics.
Own the compute, vector database, retrieval, privacy, security, and reliability architecture for the intelligence system.
Work directly with the CIO and traders to reduce time-to-insight and deliver production tools quickly.

Python with hands-on Pandas and NumPy experience is required.
Hands-on experience with LLMs, RAG architectures, prompting, evaluation, and failure modes.
Experience with vector databases such as Pinecone, Milvus, or FAISS.
Experience with orchestration tools such as LangChain, Airflow, or custom agents.
Experience with REST and WebSocket APIs for market data, news, or internal tools.
Experience building lightweight internal UIs with tools such as Streamlit, Dash, or Retool.
Understanding of production concerns including latency, hallucinations, context windows, reliability, and data privacy.
Comfort with financial data, uncertainty, discretionary judgment, and time-sensitive or adversarial environments.
Founder-minded experience building personal projects or working in a high-stakes startup, prop shop, or family office is sought.

Direct access to the CIO and decision-makers with no middle management or reporting theater.
Immediate ownership and use of systems supporting a portfolio of approximately $500M.
High autonomy and founder-level responsibility for the firm's intelligence engine.
Applicants should submit a GitHub, a project they are proud of, or a short architecture note; no cover letter is requested.

Apache AirflowNumPyPandasPython