AI/ML Engineer - Trading & AI Agent Teams - Remote from Home

Build an autonomous AI agent system that combines real-time data processing, multi-agent LLM orchestration, and advanced ML research. You'll work directly with the founder to transform a working data pipeline into an intelligent, self-improving system that analyzes markets using cutting-edge AI.

Current State:

• Real-time data ingestion operational (100K+ ticks/sec via QuestDB + Redis)
• Multi-source data feeds integrated (Databento, ThetaData, Benzinga, Unusual Whales, FRED)
• Agent orchestration framework designed (Google ADK, LangGraph)
• Basic alerting system functioning

Your Mission:

What You'll Build

Feature Engineering & Retrieval System

• Design compressed data representations for efficient LLM retrieval
• Build knowledge graph (Neo4j) linking events patterns outcomes
• Create RAG system: embed patterns, enable semantic retrieval of historical analogues
• Develop signal scoring engine combining multiple data streams

AI Agent Implementation

• Build Research Agent: LLM-driven hypothesis generator discovering patterns in data
• Implement Monitoring Agent: Real-time anomaly detection and alerting
• Create Analysis Agent: Sentiment and narrative extraction using FinBERT + LLMs
• Deploy Shadow Mode: Run agents on live data, validate against outcomes

Strategy & Automation

• Develop pattern detection using ML (matrix profile/STOMP for motif discovery)
• Build backtesting framework with walk-forward optimization
• Implement execution automation with optimization algorithms
• Create self-improvement loop: agents learn from outcomes, adapt strategies

Required Skills

Core Technical:

• Expert Python for production systems (not just notebooks)
• ML fundamentals: scikit-learn, PyTorch, embeddings, retrieval systems
• Time-series expertise: statistical analysis, pattern recognition, forecasting
• Systems design: Can architect data pipelines and production ML systems
• Data engineering: Experience with high-volume streaming data

Quantitative Analysis:

• Strong statistical analysis and hypothesis testing
• Experience with time-series data and signal processing
• Understanding of optimization algorithms and search spaces
• Ability to validate models rigorously and avoid overfitting

Preferred Skills

• LangChain/LangGraph or similar LLM orchestration frameworks
• Graph databases (Neo4j) or knowledge graph work
• Vector search (pgvector, Qdrant) and RAG architectures
• Real-time systems: Kafka, Redis Streams, QuestDB
• Deep learning: LSTM, Transformers for sequential data
• Academic background: MS/PhD in CS, Physics, Math, Statistics, Data Science

Tech Stack

Data Infrastructure:

• QuestDB (time-series DB) + Redis Streams for real-time ingestion
• PostgreSQL for analytics, Neo4j for knowledge graphs
• Multiple API integrations for diverse data sources

AI/ML Stack:

• LangChain + LangGraph for agent workflows
• Google ADK for multi-agent orchestration
• PyTorch for deep learning, scikit-learn for classical ML
• HuggingFace Transformers (sentiment, forecasting models)
• Claude, GPT. Gemini, Grok for LLM capabilities

Infrastructure:

• Python 3.11+ with async patterns
• Docker + systemd orchestration
• GitHub Actions CI/CD