Senior / Staff AI Engineer (Agent Systems)
Summary:
Multiple early-stage portfolio companies (Seed/Series A) are hiring engineers to build the next generation of AI-native systems across enterprise software, security, healthcare, developer tools, and vertical SaaS.
These companies are building systems that go beyond copilots — software that reasons, takes action, and executes real-world workflows end-to-end. While each company operates in a different domain, they share a common set of technical challenges: goal-driven execution, multi-step reasoning, system reliability, and increasingly, coordination across multiple agents and tools.
This is not a single role at one company. Applicants will be considered for multiple openings across Greylock-backed companies, and we will work directly with you to match you with the most relevant teams based on your experience and interests.
What You’ll Do:
- Design and build agent systems that execute multi-step workflows across real-world environments
- Develop goal-driven systems that plan, act, and adapt in the face of ambiguity
- Contribute to or lead the design of multi-agent architectures, where systems coordinate to complete complex tasks
- Build orchestration layers that manage tool use, decision-making, and long-running workflows
- Define and implement reliability mechanisms including evaluation, guardrails, recovery, and human-in-the-loop systems
- Drive system performance across latency, cost, and quality in production
- Partner with founders, product, and engineering teams to translate open-ended problems into working systems
- Help set technical direction and establish best practices as teams scale
Where You’ll Work in the Stack:
- Planning & Reasoning: goal decomposition, decision-making, long-horizon execution
- Orchestration: workflow management, tool routing, multi-component systems
- Execution: integrations with APIs, services, and real-world systems
- State & Memory: persistent context, user/system state, temporal reasoning
- Evaluation & Control: feedback loops, safety systems, observability
- Infrastructure: deployment, scaling, monitoring, and performance optimization
What We’re Looking For:
- Experience building and shipping production LLM or agent-based systems that take actions, not just generate outputs
- Strong understanding of goal-driven execution and multi-step workflows
- Experience with orchestration, tool use, or multi-component systems (e.g., ReAct, LangGraph, or similar approaches)
- Track record of owning systems end-to-end, including reliability and iteration in production
- Comfort working across the stack, from model behavior to system integration
- Strong engineering fundamentals (Python or similar; experience building production systems required)
- Ability to operate in ambiguous environments and make high-leverage technical decisions
Signals We’re Especially Excited About:
- You’ve built systems that complete workflows end-to-end, not just assist users
- You’ve worked on agents that operate over time (stateful, proactive, or long-horizon systems)
- You’ve designed or contributed to multi-agent or multi-component systems
- You’ve owned failure modes, reliability, and recovery in production systems
- You’ve integrated AI systems with external tools, APIs, or enterprise environments
- You’ve built from first principles in early-stage settings
About Greylock
Greylock is a leading early-stage venture capital firm that partners with exceptional founders building category-defining companies. Our portfolio includes Figma, Anthropic, Ramp, Abnormal Security, Rubrik, Airbnb, LinkedIn, Roblox, Dropbox, and Coinbase.
About the Greylock Recruiting Team
As full-time employees of Greylock, our team works across the portfolio to connect candidates with open roles at our portfolio companies. If there is a strong fit, we will introduce you directly to hiring teams for full-time opportunities at those companies.
