3 days ago
Remote, United StatesStaff+
Responsibilities
- Own the architecture, roadmap, production rollout, and on-call operation of agentic onboarding and self-healing automation systems.
- Deploy the existing API catalog and browser workflow agents using durable workflows and production observability.
- Detect drift in APIs and web interfaces, diagnose failures, repair workflows or specifications, and re-validate fixes against real runs.
- Define evaluation benchmarks and quality thresholds for generated specifications, workflows, and repairs.
- Optimize output cost and latency by selecting models based on measured accuracy and cost.
- Own threat modeling and security controls for prompt injection, credential isolation, token handling, sandboxed browsing, and egress.
- Partner with integrations, automation-runtime, and platform teams, write technical design documents, and mentor contributors.
Requirements
- Experience shipping an LLM agent with tool use that ran unattended for real users and owning a backend system in production.
- Strong Python or TypeScript skills with productivity in both.
- Experience with LLM agent loops, structured outputs, retries, and agent failure modes.
- Experience building benchmarks or evaluation harnesses and using them to make engineering decisions.
- Depth in browser automation such as Playwright or in backend and API-integration systems, with comfort in the other area.
- Experience with distributed-systems concepts including queues, idempotency, and durable workflow engines such as Temporal or Step Functions.
- Strong security judgment for systems handling secrets, including prompt-injection threats.
- Preferred experience with Temporal in production, Pydantic AI or a comparable agent framework, open-weight model serving, LLM cost modeling, computer-use agents, browser extensions, OpenAPI tooling, identity, IAM, credential management, AWS, or Terraform.
- Clear technical writing, sound trade-off judgment, and the ability to take early systems to production and keep them reliable after launch.
Tech Stack
Categories
About Cerby
Cerby builds an identity security platform for enterprise IT and security teams to secure and automate access to disconnected or nonfederated applications. Its SaaS pairs an access-management backend with a cross-browser extension for credential injection, passkeys, and lifecycle automation where SSO or SCIM are unavailable. Founded in 2020 and headquartered in Alameda, California, the company is privately held.