About the job
As a Software Engineer in GenAI Infrastructure & Agent Systems, you will design and build platforms that significantly improve developer productivity and ML research velocity. This role spans: Agent infrastructure (harnesses, sandboxes, plugin frameworks, MCP integrations) enabling reliable, production-grade AI agents; Autoresearch systems that run experiments, evaluate results, and propose improvements to core ML models. You’ll work across the stack — from cloud infrastructure and isolation to agent memory, evaluation frameworks, and developer-facing tools.
Responsibilities
Build AI agent infrastructure
Design agent orchestration, sandboxing, and isolation systems
Develop skill/plugin frameworks and integrate internal tools (e.g., CI/CD, code, observability systems)
Scale compute infrastructure for agent workloads (GCP, Kubernetes)
Develop autoresearch systems
Build self-improving agents that run experiments, analyze results, and iterate
Implement memory, feedback loops, and long-horizon reasoning
Integrate with ML training, evaluation, and data pipelines
Create AI-powered engineering tools
Build assistants for support, triage, and workflow automation
Develop tools for code generation, debugging, testing, and PR review
Detect and resolve performance and reliability issues automatically
Improve quality and developer workflows
Design evaluation and observability systems for agent performance
Integrate agents into CI/CD for testing, failure attribution, and remediation
Build knowledge systems for search, retrieval, and reuse
Qualifications
Minimum
5+ years of experience and a Bachelor’s or 4+ years of experience and a Master’s Degree in Computer Science, software engineering, a related field, or equivalent practical experience.
Strong programming skills in Python, C++, or Go, with experience building scalable, reliable systems
Solid background in cloud infrastructure (GCP/AWS), Kubernetes, CI/CD, and developer tooling
Passion for building production-grade AI agent systems and improving engineering efficiency through automation
Interest in autonomous, self-improving systems and ML-driven workflows
Strong systems thinking across infrastructure, memory, and tooling, with ability to work end-to-end
Preferred
Experience with LLM/agent systems in production (tool use, orchestration, sandboxing)
Familiarity with autonomous agent patterns (memory, reflection, planning)
Experience with MCP, plugin/skill frameworks, or similar architectures
Background in ML infrastructure (training pipelines, evaluation, experimentation)
Experience in DevEx/platform engineering, observability, or security isolation patterns