About this role
Developing the security foundations for agent execution platforms involves shaping how AI agents are sandboxed across product and internal systems at Harvey. The security engineering team builds the execution environments, orchestration, and controls that allow models with advanced cybersecurity capabilities to automatically discover and validate vulnerabilities within authorized targets.
Operating as a founding member of the security team requires leading development of a platform for security agents while balancing containment with reliability, performance, and flexibility. Engineers in this position own initiatives from design through rollout, establishing shared technical direction and partnering closely with infrastructure and product engineering teams.
What you will do
- Set the technical direction for secure agent execution by working across security, infrastructure, and product engineering teams.
- Design, build, and operate a platform enabling security agents to discover and validate vulnerabilities within authorized targets and execution boundaries.
- Build reusable controls for agent tool use, code execution, network and filesystem access, resource consumption, and workload lifecycle.
- Partner with product sandbox and internal agent platform owners to make architectural decisions, implement protections, and integrate shared security capabilities.
- Enable internal use of open-source models alongside infrastructure teams by building protections around model serving and data handling.
- Develop adversarial tests, evaluations, and telemetry that verify containment and expose failures while building mechanisms to safely terminate agent activity.
- Lead delivery and adoption across teams, mentor engineers, and remain hands-on through implementation and production operation.
What you need
- Ten or more years of experience developing and running production software, including technical leadership on initiatives spanning multiple engineering teams.
- Strong software engineering skills in languages such as Go, Rust, Python, or C++, backed by practical experience in Linux systems, networking, and containerized infrastructure.
- Deep expertise in one or more areas of execution security like sandboxing, operating-system isolation, container security, or untrusted code execution platforms.
- Experience building shared platforms, services, or libraries and driving adoption across teams through clear interfaces and safe migrations.
- Demonstrated experience with cloud infrastructure and distributed systems, covering observability, failure handling, capacity management, and dependable production operation.
- Fluency with AI-assisted engineering and agentic workflows to accelerate development and reason about risks introduced by tool-using agents.
- Strong communication and technical judgment with a record of bringing teams to agreement on ambiguous problems and carrying decisions through to delivery.
Nice to have
- Experience with microVMs, hypervisors, or sandbox runtimes such as Firecracker, gVisor, or Kata Containers, or Linux isolation mechanisms.
- Experience building agent runtimes, tool-execution frameworks, or automated security-testing platforms.
- Experience with vulnerability research, penetration testing, or adversarial evaluation of agent systems.
- Experience deploying or securing self-hosted inference and open-source models, including isolation of model-serving workloads.
Benefits
- Offers Equity
- Offers Bonus
Skills and tech stack
Questions candidates ask
What is the salary range for this role?
The compensation ranges from $231,000 to $346,400 USD. Additional offers include equity and bonus.
Where is the position located?
The role is located in San Francisco.
What kind of technologies will I work with?
The tech stack includes languages such as Go, Rust, Python, and C++, alongside Linux systems, containerized infrastructure, and sandbox runtimes like Firecracker, gVisor, or Kata Containers.
What will I be working on day to day?
You will build the security foundations and execution environments that let Harvey put increasingly capable AI agents to work. This involves leading development of a platform for agents to perform security work, sandboxing agents, and building reusable controls for tool use and code execution.