Senior/Staff Software Engineer, AI Agent Infrastructure
Core
Building a trustworthy, autonomous agent platform and closed-loop evaluation system to amplify engineering output and automate research/experiments at scale.
Role type
Staff/Principal Software Engineer (AI Agent Infrastructure & Evaluation)
Builds
Autonomous agent runtime, closed-loop evaluation systems, autoresearch infrastructure, and agent-powered tooling for the engineering lifecycle.
Domain
AI Infrastructure, Autonomous Systems, LLM Research Operations
Deliverable
production ML models | infrastructure | product features
Required skills
LLM training dynamics (SFT, RL, preference optimization), inference optimization (KV-cache, speculative decoding), distributed systems at scale, agent system architecture (orchestration, sandboxing, memory), Python, Go/C++/Rust, experiment orchestration, observability, security isolation
Preferred skills
Post-training/fine-tuning experience, developer experience (DevEx) or platform engineering, open tool-integration protocols, multi-agent coordination
Technologies
Python, Go, C++, Rust, LLM frameworks, cloud infrastructure, queuing systems
Responsibilities
Design and build the closed-loop measurement layer for agent output acceptance/reversion; automate the research loop from assisted to unattended; design isolation and permissioning models for agents acting on production systems; build agent-powered tooling for code generation, review, and debugging.
Seniority
Staff/Principal, hands-on IC with deep scope and ownership