Member of Technical Staff- Research Engineer (Harness Engineering and Agentic Orchestration)
Core
Building the context layer for AI coding agents and a platform for AI-native software development, focusing on agent harness orchestration, evaluation methodologies, and failure-mode analysis.
Role type
Research Engineer (AI Research)
Builds
AI coding agent platform, agent harnesses, evaluation infrastructure, and production pipelines for AI-native software development.
Domain
Artificial Intelligence / Software Development / B2B SaaS
Deliverable
production ML models | product features
Required skills
Agent harness and orchestration design, Agentic eval methodology, Outer-loop and pipeline thinking, Failure-mode analysis, LLMs and agentic systems experience, Dataset building and curation, Customer workflow analysis, Prototype development, Research rigour.
Preferred skills
Masters or PhD in computational field, Coding agents or code-generation systems experience, RL or bandits experience, Synthetic data infrastructure, GitHub project portfolio.
Technologies
LLMs, Agent frameworks, Evaluation datasets, Trace analysis tools, Python (implied by engineering context), GitHub.
Responsibilities
Design agent harnesses and orchestration logic, Build and maintain evaluation infrastructure and datasets, Analyze agent traces to diagnose failure modes, Prototype new harness configurations, Collaborate with engineering on near-term shipping problems, Translate research papers into concrete experiments.
Seniority
Mid-Senior, hands-on IC