About the Product
QuantPilot is 3Commas' AI trading copilot. Its agents research markets, generate trading strategies, backtest them, and iterate on them with traders in chat. The product provides institutional-grade research tools to serious retail traders.
Responsibilities
•Own the agent architecture end to end, including orchestration, tool loops, context management, sandboxes, model routing, and production quality, cost, and latency trade-offs
•Own the contracts between the engine and the product, including how the frontend renders agent events
•Build and own the evaluation layer, including offline evaluations, backtest-based quality metrics, and regression detection for prompt or model changes
•Apply statistical rigor to strategy evaluation through overfitting detection, robustness checks, and walk-forward and out-of-sample validation
•Build the data foundations the agent reasons over, including provenance, validation, and a consistent model of market entities
•Make the system observable enough to diagnose non-deterministic production failures
•Refine requirements with product, set technical direction through RFCs and ADRs, and mentor engineers moving into LLM systems
•Participate in the on-call rotation for owned services, respond to production incidents, and follow the incident response process
Requirements
•5+ years of backend or ML engineering, including at least one year building LLM-based systems in production
•Understand agent failures below the framework level and be able to fix them in your own code
•Demonstrated eval-driven development, including maintaining a quality metric over time and using it to improve an agent
•Strong Python and production engineering fundamentals, including services, queues, streaming, observability, and Kubernetes
•Be comfortable following a bug into TypeScript or Go
•Apply statistical principles to backtesting methodology
•Have a product mindset and turn ambiguous goals into shipped, measured results
•Working proficiency in English (B2+)
•Use AI-assisted development tools responsibly and remain accountable for the correctness, security, maintainability, and production impact of shipped work
Nice to Have
•Quant finance background, backtesting frameworks, market data and strategy research
•Knowledge graphs, ontologies, semantic layers, RAG
•Go
•Software built on the receiving end of the MCP protocol
•High-load low-latency systems in fintech or trading
•Language model fine-tuning or training
Conditions
No conditions stated in the posting.
