Applied AI Engineer, Kernel Performance

Etched • San Jose, CA

20h • $150,000 - $225,000 • Onsite

About The Position

Etched is building hardware for frontier intelligence, focusing on co-designing chips, racks, software, and manufacturing to deliver best-in-class throughput and latency for inference workloads. The company is backed by significant investment and staffed by leading engineers. This role involves building AI systems that autonomously transform newly released model architectures into correct, production-ready implementations optimized for Etched hardware. These systems will explore broader design spaces, learn from experiments, and achieve peak performance faster than traditional methods. Etched offers a unique research loop with proprietary hardware, compiler, runtime, kernels, production workloads, and dedicated compute resources, allowing for teaching models with proprietary performance signals and iterating on proposals to improve both the optimization system and the hardware.

Requirements

• A track record of solving hard problems across stacks and domains — you enjoy being dropped into unfamiliar territory and figuring it out
• Comfort with both Python and low-level code: you can read it, modify it, debug it, and direct AI to write it well. We do not care whether you write code from scratch — we care whether you ship things that work.
• Kernel experience: you've written or tuned kernels and can explain the mechanisms and performance impact of optimizations you’ve shipped
• Fluency using AI to learn and ramp on new problems — agentic coding tools, deep research, and frontier models are how you work, not an add-on
• Moving fluidly between research exploration, agentic experimentation, low-level debugging, and production execution.

Nice To Haves

• First principles thinking on accelerator performance: memory hierarchy, data movement, parallelism, synchronization, and low-precision computation.
• Hands-on experience building and shipping LLM-based agents or AI tooling that real users depend on in production environments (beyond calling an API — context engineering, tool integration, orchestration, failure analysis)
• An eval-driven mindset: you measure whether AI systems work before scaling them
• Fine-tuning or post-training, RAG over proprietary data, and/or multi-agent orchestration
• High agency and comfort with ambiguity — you find the real problem to solve

Responsibilities

• Own the system that turns new model architectures into verified, production-ready kernels and model mappings.
• Build agents that understand Etched hardware, design experiments, generate implementations, compile and profile them, diagnose bottlenecks, and iterate with our teams, to the limits of model autonomy.
• Design evals covering correctness, numerical stability, latency and efficiency.
• Turn profiler traces, simulation, hardware counters, and expert judgment into structured signals models can learn from.
• Curate proprietary datasets and optimization memory from complete trajectories, expert demonstrations, counterexamples, and production outcomes.
• Build fast, reproducible experiment infrastructure and observability so experiments remain interpretable, trustworthy, and high-throughput.
• Ship model-generated improvements to production and quantify their impact on end-to-end system performance.
• Partner deeply with other architecture teams to shape new abstractions and Etched’s hardware-software roadmap.
• Continuously evaluate new model releases and deploy the best for each stage of the optimization loop.

Benefits

• Medical, dental, and vision packages with generous premium coverage
• $500 per month credit for waiving medical benefits
• Housing subsidy of $2k per month for those living within walking distance of the office
• Relocation support for those moving to San Jose (Santana Row)
• Various wellness benefits covering fitness, mental health, and more
• Daily lunch and dinner in our office
• Unlimited compute budget subject to ROI justification

What This Job Offers

• Job Type: Full-time
• Career Level: Senior
• Education Level: No Education Listed