About Anthropic

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.

About the role

The Claude Partner Network is how partners build, sell, and deliver on Claude. Partner Support keeps them moving: getting set up, getting certified, registering their work, and understanding where they stand in the program. It runs in layers. Self-serve and an AI support agent handle what they can, an outsourced vendor team works the in-scope (L1) topics, and our specialists take the escalated (L2) cases that need judgment.

As our Partner Support Operations Lead, you will run partner support end to end: the team, the operating model, and the systems behind every layer. Partner Programs sets program policy. You will build and run the service levels, knowledge, quality bar, and escalation paths that put that policy into practice, and keep them reliable as volume grows.

This is a hands-on operator and builder role. You will be close enough to the queue to know why a case was hard or a rule broke, while you build the team, the vendor coverage, and the Claude-powered automation that keep support from growing in line with partner count.

Key responsibilities

Manage the partner support team: set priorities and clear expectations, coach the Partner Support Specialists, address performance early, and expand the team if volume or coverage calls for it
Own the support operating model: the L1 and L2 boundary, severity definitions, service levels, and escalation paths; review misrouted and wrongly escalated tickets, and fix the rule as well as the ticket
Run the L2 queue to its service levels: coverage across shifts, one owner for every case, clean handoffs to the teams that hold the answer, and careful handling of sensitive and high-impact partner issues
Manage the support service line with our business process outsourcing (BPO) team: work with their leads day to day and hold a stated quality bar through training and sign-off for each new cohort, the QA rubric and calibration, the weekly operating review, and corrective action when quality or coverage slips
Own the knowledge layer: one source of truth for the help center, macros, and AI agent content, updated when a program change ships, so every layer gives partners the same answer from day one of a launch
Own the support platform: routing, conversation attributes, views, workflows, access, and service-level reporting; keep the taxonomy short, consistent, and reportable
Scale through self-serve, vendors, and Claude: use topic and outcome data to decide what moves to each, and build automations for tagging, routing, QA screening, and reporting; test them on real tickets and add human checkpoints where a wrong answer carries risk
Turn what partners ask into fixes: bring the recurring causes behind ticket volume to Partner Programs, Partner Business Systems, and other owners, and follow each one through until the tickets stop
Run support on a scorecard covering responsiveness, resolution, escalation accuracy, and quality, and use it to plan capacity and to tell the team, the vendor, and leadership what to fix next

Minimum qualifications

Support operations experience at a technology company, including managing a small team
Experience running an outsourced or BPO support team: onboarding waves, service levels, QA rubrics, calibration, and performance feedback
A track record of building a tiered support model from an early stage and running it as volume grew
Hands-on depth in Intercom or a comparable platform (Zendesk, Salesforce Service Cloud): routing rules, attributes, workflows, reporting, and permissions
Experience running knowledge and content for an AI support agent, including how to review answer quality and close content gaps
Strong analytical instincts: able to read ticket data, find the cause behind a spike or a backlog, and turn it into a fix and a capacity plan