M
AI Engineering Lead
On-siteFull-timeAI EngineeringCybersecurity
Own the standards and infrastructure for production AI agents across platform engineering, evaluation, observability, cost control, and company-wide enablement. This is a hands-on lead role without direct reports.
Responsibilities
- Define and build a shared agent framework, including components, orchestration patterns, tool interfaces, and engineering conventions.
- Build CI-based evaluation gates using trajectory testing, golden datasets, offline replay, and per-dimension scoring.
- Own agent observability across execution graphs, tool calls, intermediate reasoning, latency, and quality drift.
- Improve observability for externally orchestrated workflows, including Temporal-based flows.
- Manage agent economics through provider abstraction, model routing, model substitution, and per-flow cost attribution.
- Improve production agent latency and determinism through tighter loops, early stopping, and deterministic state machines.
- Create internal AI tooling, skills, and playbooks, and teach teams when and how to use AI effectively.
- Partner with security research, product, and engineering leadership to align agent capabilities with the product roadmap.
- Publish methodology for benchmarking agentic security output and modeling residual risk.
Requirements
- Production experience shipping agentic systems with orchestration, tool use, structured outputs, and large-scale failure handling.
- Experience building evaluation systems with trajectory tests, golden datasets, regression gates, and offline replay.
- Strong backend and distributed-systems engineering skills.
- Strong Python skills and experience with Temporal or equivalent workflow orchestration, streaming, and cloud-native infrastructure.
- Fluency with agent frameworks, multi-provider model routing, structured output contracts, prompt engineering, and context engineering.
- Security literacy sufficient to assess the trustworthiness of AI security outputs.
- Demonstrated ability to influence multiple teams without direct authority.
Nice to have
- Experience with AI/LLM security, including agent red-teaming, prompt injection, or agentic attack patterns.
- Background in cybersecurity, detection engineering, or WAF or mitigation systems.
- Experience driving AI adoption across an entire company.