AI Research Engineer

Stack

Responsibilities

• ▹Design and implement multi-agent and RL approaches for agentic code generation and tool use
• ▹Build research prototypes that integrate with the platform; collaborate to productionize wins
• ▹Create evaluation suites: task specs, pass/fail checkers, coverage, cost/latency dashboards
• ▹Acquire and curate datasets from complex technical documents such as chip specifications; generate synthetic data where appropriate
• ▹Analyze experiments with disciplined ablations; document results and decisions
• ▹Stay current on LLM agents, RL (offline/online, RLHF/RLAIF), constrained decoding, and program synthesis

Requirements

• ▹PhD in CS/AI/ML (or equivalent research experience), ideally with publications in multi-agent RL, agentic AI, or RL for language/code
• ▹Strong Python and ML framework experience (PyTorch preferred; JAX/HF a plus)
• ▹Demonstrated ability to turn research into working systems; reproducibility mindset (tests, seeds, configs, logging)
• ▹Experience designing eval harnesses and success metrics for sequential/agentic tasks
• ▹Comfortable with data acquisition and curation from documents and logs, with good instincts about data quality and licenses

Nice to have

• ▹Research on program synthesis/codegen, constrained decoding, or execution-based rewards
• ▹Experience with offline RL from tool traces or human corrections
• ▹Open-source contributions (e.g. CleanRL, RLlib, AutoGen, LangGraph, CrewAI, Transformers)
• ▹Familiarity with semiconductor/chip domains or other complex technical specs
• ▹Track record of shipping research to production and measuring impact

Soft skills

About the company

Normal Computing builds silicon that turns thermal noise from an obstacle into a computational resource — stochastic, in-memory, asynchronous architectures that deliver 10-100x more AI inference per dollar and per watt. The company co-designs the full stack, from AI-native EDA systems used by the world's largest semiconductor companies to the advanced ASICs they make possible. Backed by $85M+ from leading deep-tech investors, the team works across New York, Silicon Valley, London, Copenhagen, and Seoul.