Head of Technical Support

Location: San Francisco, in person (5 days/week)

Team: Technical Customer Success / Support

Reports to: CTO

LiteLLM is the open-source AI gateway powering production AI traffic for NASA, Adobe, Netflix, Stripe, Nvidia, and many other Fortune 500s. We provide a unified API across 100+ LLM providers and act as the enterprise control plane for AI traffic. We're a YC W23 company, post-Series A, with 36K+ GitHub stars and growing fast.

As we scale, our enterprise customers expect support that matches the criticality of the workloads they run on us. We need someone to own that.

You will own the support function end to end and drive escalations to zero. Specifically:

1. Build our support processes. When something escalates today, it skips three layers and hits our founders. You'll build the layers that should have caught it.

2. Build the system that gets us to zero dropped threads. Define and enforce SLAs across every channel (Slack Connect, email, Zendesk). Every customer thread has an owner, an ETA, and a clear next step at all times.

3. Make support legible. Set up performance tracking for the support team. Tell us with data whether we're under-resourced, whether someone is overloaded, or whether someone is underperforming. We can't fix what we can't see.

4. Hold engineering accountable for customer tickets. Today, customer-reported bugs sit in engineering queues with no clear owner or ETA. You will build the system that closes that loop, weekly.

5. Own your own hiring. You'll tell us when we need more support engineers vs. more managers vs. better tooling, and you'll run that hiring yourself.

• Audit every active customer thread across Slack, email, and Zendesk. Identify the dropped ones.

• Define SLAs by customer tier and by issue severity. Get team and customer buy-in.

• Stand up a weekly support metrics review: response time, resolution time, escalation rate, ticket aging, NPS by account.

• Build the engineering accountability loop: every customer-reported issue has an owner, an ETA, and a weekly review with engineering leadership.

• Diagnose whether current escalations are systemic (product/quality issues), team capacity issues, or process gaps. Bring us the data.

• Make the call on next hires: support engineer, support manager, or TAM.

• Obsessive about inbox zero and Slack zero. You physically cannot leave a thread without an owner and a next step. You feel it.

• Systems-driven. You don't just close tickets, you have experience designing the system that closes tickets without you. You've built support orgs (or comparable ops functions) from 2 to 10+ people.

• Comfortable evaluating people with data. You can look at your team's load and output and tell us, with metrics, whether the team needs more headcount or more coaching.

• Knows where your role starts and ends. When a ticket comes in, you ask "is this systemic or specific to this account?" before reacting. You don't try to debug the product yourself, but you know enough to challenge engineering on ETAs.

• Customer-credible. You can run a tense customer call without overcommitting on a timeline you can't hit. You've been the person on the other end of a Fortune 500 escalation and held the line.

• Comfortable with ambiguity and speed. This is a startup. The system you build in month one will be wrong by month three, and you'll rebuild it.

Bonus:

• Background supporting developer-facing or infrastructure products

• Experience with Zendesk, Linear, and Slack Connect at scale

• Familiarity with LLMs, AI infrastructure, or the gateway/proxy space

• $250,000 to $300,000 base salary plus bonus and equity and benefits. We pay at or above market for the right person, and the equity is meaningful at our stage and growth. We also have 401(k) with a 3.5% match of employee contributions.