SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Anthropic is hiring an AI Operations Engineer to design, build, and run agentic workflows that automate partner experience operations. You'll work within the Partner Experience team, which is building an AI-native operating model on top of existing systems: a partner portal, CRM workflows, enablement and certification platforms, and data integrations.
In this builder role, you'll own the full lifecycle of Claude-powered agents and automations. Key responsibilities include:
• Design and build agentic workflows using Claude, MCP connectors, and API integrations to automate partner application triage, onboarding, deal registration, certification tracking, communications, and reporting.
• Integrate these workflows with the partner portal, CRM, enablement stack, finance systems, and databases—consuming and writing back to systems of record rather than creating shadow processes.
• Convert recurring manual work into reliable, monitored automations with guardrails, human-in-the-loop checkpoints where decisions carry risk, and comprehensive documentation.
• Own operational health: error handling, edge cases, escalation paths, and iteration as partner processes evolve.
• Prototype rapidly with the team—identify high-friction workflows and deliver working v1s in days, not quarters.
• Establish reusable patterns (skills, prompts, connectors, playbooks) to accelerate future builds and enable broader team self-service.
You'll ship working automations weekly and continuously expand team capacity without adding headcount. This is a hands-on engineering role focused on practical impact.
Required qualifications: hands-on experience building with LLMs and agentic tooling (prompt design, tool/function calling, MCP or comparable frameworks, agent output evaluation); experience integrating across APIs and business systems (Salesforce, partner portals, LMS platforms, Slack, email) using Python or JavaScript; strong judgment about automation risk and human-in-the-loop design; and the ability to move fast and iterate based on real partner feedback.
