AI Research Resident

The Residency Program

The AI Research Residency is Normal Computing's flagship program for exceptional researchers and engineers who want to work at the frontier of AI. Residents join a small, hand-picked cohort with a dedicated research mentor, direct access to the teams building our in-memory inference hardware and the AI systems that design and verify it, and a clear arc from onboarding through publication.

Every residency is built around two milestones that mark you as part of something bigger than a single project: a research paper co-authored with our team, and a company-wide research colloquium where you present your findings to the full Normal Computing organization. You'll leave with a body of published, presented work, and a standing as one of the earliest residents to help define what this program becomes.

This Track: Autoresearch

Normal's hardware runs the most memory-bound parts of large-model inference inside memory itself, rather than shuttling weights to a processor. Building silicon like that quickly needs AI in the loop: agentic systems that explore the design space, generate and verify candidate designs, and learn from every experiment, in simulation, in the lab and on deployed silicon. This track is about that loop. You will use AI to make our hardware and the tools that design it better, and use what better hardware teaches us to make the AI better in turn, a closed loop where each side accelerates the other.

You'll treat recursive self-improvement as an empirical engineering problem: build the strategies, measure whether each improvement actually compounds, and feed the result back into the next generation of chip.

Your Normal Experience

This is a deliberately cross-functional residency, one of the few seats at Normal with a foot in both worlds. On one side is our thermodynamic hardware: silicon that exploits noise and analog dynamics rather than fighting them. On the other is our autoresearch platform: agentic systems that learn from design data to design and verify that silicon. You'll work across both, looking for the places where an improvement to one compounds into an improvement to the other, and turn research into production-quality research code that a team can build on.

You'll get direct exposure to the pace, ambiguity, and speed of decision-making that comes with working at a fast-moving, well-funded hardware startup, where the distance between an idea on a whiteboard and a test on real silicon is measured in weeks, not years.

What You'll Do

• Develop recursive self-improvement strategies. Design and test strategies where AI systems improve the tools, workflows, and hardware used to build the next generation of AI systems, closing the loop between design tools, silicon, and the models that run on it.
• Apply AI to accelerate hardware design. Work on agentic and ML-driven approaches to chip design and verification, looking for places where automation and learned models compress design cycles that used to take months into weeks.
• Use hardware advances to accelerate AI itself. Take what's learned from thermodynamic and unconventional hardware and feed it back into how we train, evaluate, and iterate on our own AI tools and research.
• Build the evaluation infrastructure this loop needs. Help build benchmarks and evaluation frameworks that can tell, rigorously, whether a given self-improvement actually compounds. This is empirical work, not a leap of faith.
• Drive a research question of your own. Partner with the hardware and AI research teams to scope, run, and iterate on an original technical investigation that lives at their intersection.
• Co-author a paper. Work with the team to write up your findings for submission to a relevant venue, with mentorship on framing, experiments, and technical writing. One of the two milestones every resident builds toward.
• Present at the residency colloquium. Share your work and thinking with the broader Normal Computing research community at the program's capstone event, and get real-time feedback from people building this technology every day.
• Experience startup pace firsthand. Work directly with founders, senior researchers, and engineers in a lean, fast-moving environment where priorities shift quickly and your work has an outsized, visible impact.

What Would Make You a Great Fit

• Currently pursuing or recently completed a graduate degree (MS or PhD, or equivalent research experience) in computer science, AI, electrical engineering, physics, or a related field. Publications are a strong plus, but not required at the resident level; a solid research track record is what matters.
• Cross-functional by temperament: comfortable moving between hardware and software, physics and algorithms, and not content to stay on one side of that line.
• Strong Python skills and proficiency with modern ML frameworks (PyTorch preferred), plus some exposure to hardware-side tooling (SPICE, FPGA or ASIC flows, or EDA software) or a strong eagerness to pick it up.
• Familiarity with agentic systems, reinforcement learning, or automated design and optimization loops: the kinds of methods that let one system meaningfully improve another.
• Systems-level thinking: you can reason about second-order effects, where an improvement in one layer (say, a design tool) changes what's possible in another (say, chip performance, or model quality).
• A rigorous, evidence-first mindset about self-improvement. You're drawn to this problem because it's a hard empirical and engineering challenge, and you care about measuring real gains rather than assuming them.
• Strong written and verbal communication skills; prior experience writing up research (papers, theses, technical reports) is a plus, as is any experience presenting technical work to a live audience.
• A bias toward ownership and self-direction. You're energized, not overwhelmed, by the ambiguity and speed of a small, fast-moving startup.