Internal Acceleration Research Engineer (Kernels)

Salary: $150,000 - 400,000 per year Requirements:

• 1–5+ years of experience building and shipping software end to end, ideally at a fast-moving startup.
• Experience building agents for complex, multi-step tasks that people rely on, including tool design, long-context management, evaluations, and failure analysis.
• Strong software engineering skills, with experience building reliable systems such as distributed systems, experimentation infrastructure, or developer tools.
• Heavy use of coding agents, including reviewing their transcripts and improving systems when agents produce confidently incorrect results.
• Ability to work independently in unfamiliar domains, identify the underlying problem, and drive solutions forward.
• Product sense, design judgment, and exceptional attention to detail.
• Helpful additional experience includes chip development, kernels, hardware, manufacturing, infrastructure, data engineering, DevOps, fine-tuning, RAG over proprietary technical data, multi-agent orchestration, or startup hypergrowth.

• Work with our kernels team to identify high-impact bottlenecks and take solutions from initial concept to tools engineers rely on every day.
• Build and deploy LLM agents integrated into our engineering workflows through tool-calling and MCP.
• Create execution environments where agents can verify their work using sandboxed simulators, EDA flows, compilers, profilers, and real hardware.
• Develop full-stack internal products from the ground up, including interfaces, data pipelines, and infrastructure.
• Design evaluations and monitor adoption to determine which tools to expand and which to discontinue.
• Take on special projects with our founders, identify needs across the company, and help deliver critical milestones.

• AI
• DevOps
• Fine-tuning
• Hardware
• Support
• LLM
• MCP
• RAG
• API

More:

Our Internal Acceleration team builds agents to accelerate the production of frontier tokens. In this role, youll work directly with our founders and alongside our kernels team to develop agentic harnesses and operating systems for building hardware in an AI-native world, covering tools, context and memory, execution environments, and evaluations. Were looking for engineers who enjoy ambiguity, take end-to-end ownership, and are energized by time-sensitive, business-critical work; this is not a 9-to-5 role. Etched builds hardware for frontier intelligence by co-designing chips, racks, software, and manufacturing, with an initial focus on inference. We believe in the Bitter Lesson and are building an inference-focused frontier AI system. Our team is fully in person in San Joses Santana Row, and technical staff contribute across engineering and research. We offer medical, dental, and vision coverage, a $500 monthly credit for waiving medical benefits, a $2,500 monthly housing subsidy for employees within walking distance of the office, relocation support to San Jose, wellness benefits, daily office lunch and dinner, and a compute budget subject to ROI justification. Base compensation is $150,000–$400,000+, plus significant equity.

last updated 42 week of 2026