Research Engineer

Research Engineer — AI Agents & Hardware We’re partnering with an ambitious, well-funded stealth startup building at the intersection of AI agents and silicon. The team is developing agentic systems that can reason about complex engineering problems, interact with real tools and accelerate how hardware is designed and verified. This is an early, high-impact hire for an engineer who enjoys moving between research and implementation. You’ll help develop agent environments, training data and evaluation infrastructure, while also shipping the product and platform capabilities needed to put that research into practice. What you’ll work on Research and build AI agents for complex technical workflows Create environments, benchmarks and evaluations for agent capabilities Generate and curate synthetic data, including tasks for reinforcement learning Turn promising research ideas into reliable production systems Work closely with hardware, infrastructure and product engineers Influence the technical direction of an early-stage company What we’re looking for Excellent software-engineering fundamentals Significant experience using coding agents such as Cursor or Claude Code Notable AI-agent research projects and a strong implementation projects An understanding of digital design or VLSI, gained through coursework, personal projects or industry experience Experience with areas such as FPGA development, EDA tooling, design verification or hardware-support software is particularly relevant A practical, low-ego approach and willingness to contribute across research, product and infrastructure Particularly valuable Experience at a frontier AI lab or highly regarded AI research startup Post-training data selection or filtering work Synthetic data generation for AI agents or reinforcement learning Development of agent benchmarks or RL environments You’ll join a small, highly technical team tackling a genuinely difficult AI-and-hardware problem, with substantial ownership from the outset. This is a rare opportunity to shape both the research agenda and the product built around it. Show more Show less