The Role:

You own the architecture and engineering of the ITSM platform's agentic IT operations SaaS product. You lead 120+ engineers across component teams.

You set the architecture, guide each team's design, and integrate the components into one enterprise product. Program managers own delivery tracking; you own technical readiness. You stay hands-on, debugging, writing code, and running the AI-assisted engineering harness yourself.

What You'll Do:

- Own Tensai's end-to-end architecture: agents, orchestration, the knowledge and retrieval layer, integrations, security, and scale.

- Guide each team on architecting its component and integrate the components into one enterprise solution.

- Validate the teams' work through design reviews, code reviews, and quality gates.

- Partner with program managers to flag technical risks and cross-team dependencies early.

- Build and run the team's Claude Code harness: skills, subagents, MCP integrations, test and security gates, and CI.

- Set model strategy and token economics: model choice, routing, caching, and cost per task.

- Direct SLM fine-tuning where it pays off.

- Debug hard problems and write code yourself when needed.

Must Have:

- Designed and built autonomous agents running in production: planning, tool use, multi-agent orchestration, guardrails, human-in-the-loop controls, and evaluation.

- Built knowledge systems that produce grounded, cited answers: taxonomy and ontology design, semantic modeling, knowledge graphs, ontology-grounded RAG, and retrieval that preserves global document context.

- Built and shipped SaaS products end to end at enterprise scale.

- Deep IT operations background: ITSM, observability, AIOps, ServiceNow, or similar.

- Hands-on mastery of Claude Code or a similar agentic coding harness.

- Strong coding and debugging in Python. Cloud, Kubernetes, CI/CD, and security by design.

- Led large engineering teams through architects and managers.

Nice to Have:

- SOC 2 and enterprise customer security reviews.

- Hands-on SLM fine-tuning and GPU inference.