AI Operations Engineer, Partnerships

Anthropic’s mission is to create reliable, interpretable, and steerable AI systems. We want AI to be safe and beneficial for our users and for society as a whole. Our team is a quickly growing group of committed researchers, engineers, policy experts, and business leaders working together to build beneficial AI systems.

Our Partner Experience team is building an AI-native operating model. We already run a growing set of partner-facing systems: a partner portal, CRM workflows, enablement and certification platforms, and the trackers that connect them. The systems lead on our team owns that foundation — the portal, the data model, and the integrations that keep partner records accurate and flowing.

This role is the other half of that build. As our AI Operations Engineer, you'll design, build, and run the agentic workflows that sit on top of those systems: Claude-powered agents and automations that triage partner applications, draft and send partner communications, sync data across tools, surface deal and enablement signals, and take routine operational work off the team entirely. Where the systems lead builds the rails, you build what runs on them.

This is a builder role, not an analyst role. You'll ship working automations weekly, own them in production, and continuously expand what the team can do without adding headcount.

• Design and build agentic workflows (using Claude, MCP connectors, and API integrations) that automate core partner experience operations: application triage, onboarding, deal registration, certification tracking, partner communications, and reporting

• Partner closely with our systems lead to integrate these workflows with the partner portal, CRM, enablement stack, finance systems, and database systems — consuming and writing back to the systems of record rather than creating shadow processes

• Turn recurring manual work into reliable, monitored automations: define the workflow, build it, test it against real cases, add guardrails and human-in-the-loop checkpoints where decisions carry risk, and document how it runs

• Own the operational health of what you ship: error handling, edge cases, escalation paths, and iteration when partner processes change

• Prototype fast with the team: sit with partner managers, find the highest-friction workflows, and get a working v1 in front of them in days, not quarters

• Establish reusable patterns (skills, prompts, connectors, playbooks) so each new