Research Engineer: Agentic AI for Silicon Design

Drive Capital; København; On-site; DKK 900,000 - 1,300,000; Full time

Job summary

Normal Computing seeks a Research Engineer to push the frontier of agentic LLMs and reinforcement learning for its AI platform in semiconductor design automation. You will design experiments, build agents, curate datasets, and write production-quality research code, collaborating closely with engineering to ship improvements to customers.

You’ll work on multi-agent systems, RL environments, and evaluation frameworks, staying current with advances in constrained decoding and program synthesis to

Qualifications

• PhD in CS/AI/ML (or equivalent research experience) with publications.
• Strong Python and ML framework experience (PyTorch preferred; JAX/HF a plus).
• Experience turning research into working systems and designing evaluation environments.
• Familiarity with EDA tools (formal, simulation, physical design).

Responsibilities

• Build multi-agent systems for code generation that interact with EDA tools, propose fixes, and iterate through all stages of chip design and verification flows.
• Build research prototypes that integrate with production agentic code generation tool; collaborate to productionize wins.
• Create RL environments and evaluations for agents; explore proxy rewards and speed/accuracy tradeoffs of custom tools.
• Generate datasets from silicon collateral sources (RTL, testbenches, VIPs) and licenses; synthesize data where appropriate.
• Analyze experiments with ablations; document results and drive progress with technical rigor.
• Stay current on LLM agents, RL (offline/online), RLHF/RLAIF, constrained decoding, and program synthesis.

Skills

Python

ML frameworks

Multi-agent RL

Agentic AI

Experimental design

Education

PhD in CS/AI/ML or equivalent

Tools

PyTorch

JAX

HuggingFace

Job description

Normal Computing seeks a Research Engineer to push the frontier of agentic LLMs and reinforcement learning for its AI platform in semiconductor design automation. You will design experiments, build agents, curate datasets, and write production-quality research code, collaborating closely with engineering to ship improvements to customers.

You’ll work on multi-agent systems, RL environments, and evaluation frameworks, staying current with advances in constrained decoding and program synthesis to