AI research company is building an agent-based autonomous research and execution platform for financial markets with a long-term vision across life sciences, biotech, and physics.
Responsibilities:
- Design and implement a multi-agent orchestration framework (roles, communication, memory, decision loops)
- Build retrieval and knowledge systems grounded in market data and research
- Develop a Strategy DSL and compiler for research → simulation → production
- Create falsification-first evaluation systems to eliminate false alpha
- Design an "alpha memory" layer to accumulate knowledge and avoid repeated mistakes
- Work with researchers and engineers to bring ideas into low-latency production
- Define robustness metrics beyond P&L (stability, execution realism, capacity, novelty)
Requirements:
- LLM architecture expertise: RAG, fine-tuning, prompt engineering, evaluation frameworks
- Agent systems experience: multi-agent orchestration, memory management, tool use, collaboration
- Proven experience building autonomous research/ co-scientist systems
- Strong Python and ML skills; production-ready code with PyTorch, JAX, or similar
- Statistical rigor: experimental design and statistics for non-stationary, noisy environments
- Systems thinking: design of abstractions, interfaces, and pipelines
- English proficiency: B2+
Nice to have: HFT/MFT experience or low-latency, hardware-aware, deterministic design.
Conditions & Benefits: Relocation support to Amsterdam or Dubai, or official employment in London or Montreal if already based there; access to alternative high-performance computing; research-driven environment; small fast-moving team.