Member of the Technical Staff, Software Engineering

Description

Coding agents now ship software faster than people can review or secure it. Teams see over 100 pull requests per engineer each week, and attackers use the same agents to find and exploit weaknesses within days rather than months. Quarterly pentests and endless alert queues cannot keep pace.

The company runs AI security agents over a live graph of a customer's code, CI, cloud, and telemetry. The agents attack a faithful copy of that system the way a real adversary would. Each finding arrives validated, with its blast radius and a fix plan that goes into the customer's workflow for a coding agent to apply. Integration takes hours, and the first validated finding arrives within days.

The founders previously built a security product that Datadog acquired, and security tools that thousands of organizations relied on. The company is backed by Accel and Y Combinator and is at an early stage.

Responsibilities

• Build the system that makes the agents effective: the graph they reason over, the infrastructure they run on, and the agents themselves.
• Expect no fixed role boundaries. One week could cover a new cloud integration, a change to how the graph links an endpoint to the code serving it, and tuning an offensive agent to chain two low-severity findings into a working exploit.
• Own problems from start to finish: speak with customers, design the fix, ship it, and monitor it in production.
• Work alongside the founders and a small group of engineers who deploy to production daily and use coding agents as heavily as the customers do.
• Design, test, and improve agents that pentest, triage, chain, and remediate findings, and write the evals that show whether they improve.
• Extend the live graph with crawlers that find cloud, SaaS, and code resources, and the pipeline that builds a consistent model of each production system.
• Run services in Go, TypeScript, PostgreSQL, graph databases, and AWS, from design through deployment and on-call.
• Harden the platform for production access, including strict tenant isolation, egress controls, and secrets handling that customers' security teams will review closely.
• Work with customers and security leads to turn real findings, misses, and false positives into product changes.
• Help set engineering standards as the team grows, covering code quality, review practices, tooling, and how the team uses AI agents.

Requirements

• Solid software engineering fundamentals and a history of shipping and owning production systems end to end.
• Proficiency in Go or TypeScript, and willingness to work across the stack.
• Experience with distributed systems, data pipelines, or cloud infrastructure on AWS, GCP, or Azure.
• You use AI coding agents such as Claude Code or Cursor daily and have views on making them reliable.
• You want to understand how attacks work, not only the alert they trigger.
• You are comfortable defining how something should work rather than following an existing process.
• Offensive security background in penetration testing, red teaming, bug bounty, or CTFs.
• Production experience building LLM agents, evals, or agent tooling.
• Experience with graph data models or graph databases such as Neptune, Neo4j, or Memgraph.
• Background in application or cloud security tools, including SAST, DAST, CSPM, or SIEM.
• Prior work at an early-stage startup or as a founding engineer.

Conditions

Base salary: [PAY RANGE], based on level and experience.

San Francisco based, working in person.

Applicants from all backgrounds are welcome, including those who do not match every item above, as the company is an equal opportunity employer.

Benefits

• Early-stage equity, so you share in what the company builds.
• Medical, dental, and vision coverage.
• The equipment you need to do your best work.