How we hire at Liberate.
No black boxes. We build AI that takes the mystery out of insurance, so it would be a little rich to make our own hiring process a mystery. This is your guide to exactly how we hire — stage by stage, role by role — so you can prepare with confidence and show up as your best self.
The one thing we keep confidential is the specific questions inside each interview; that’s how we make sure everyone gets the same fair, consistent shot. Everything else is right here. And remember: a great interview goes both ways, so this is your chance to get to know us, too.
Our promise on timing: we read every application within 3 business days, and after your final interview you’ll hear back within 2 — a clear yes, no, or a specific why. Silence isn’t an answer, and we don’t do it.
Agent Engineering
You’ll build the systems our AI agents run on — the reasoning, the scale, the reliability customers trust. Here’s how the process works: four conversations, then a decision the whole panel makes together.
At a glance
- Preliminary screen - 30 min, video with the hiring manager
- Technical / problem solving - 60 min, live, shared coding environment
- Systems design - 60 min, live design exercise
- Behavioral - 45 min, live conversation
Step 1 — Application Review
A real person reads every single application - no keyword filter, no black hole. If it looks like there’s a fit, we’ll reach out about next steps. And if the timing isn’t right, we’ll still tell you. Either way, you won’t be left wondering.
Step 2 — Recruiter Screen
30 minutes · phone
A conversation about your story so far - the work you’ve owned, your agentic experience, and what’s pulling you toward Liberate. We’ll leave plenty of room for your questions, and cover the practical things together, like the role and timeline.
Step 3 — Preliminary Screen
30 minutes · video call with the hiring manager
A conversation anchored on one or two systems you’ve owned — we’d rather go deep on real examples than have you context-switch across a dozen. Expect questions about impact at scale, a hard problem you debugged, how you measured success, and a time you disagreed with a stakeholder. No direct experience with agents or LLM systems? That’s genuinely fine — we’re just as interested in how you reason about something new.
Step 4 — Technical / Problem Solving
60 minutes · shared coding environment, with an engineer from the team
One or two problems, worked through in pseudocode rather than a specific language - we care more about your reasoning, structure, and communication than syntax. Expect to talk through trade-offs, edge cases, and how you’d test your solution. No AI coding assistants in this round; we want to see how you think.
Step 5 — Systems Design
60 minutes · shared environment, with a platform engineer
An open-ended design problem that’s deliberately thin on detail at the start - the questions you ask matter as much as the design itself. We’re looking at how you handle ambiguity, make and defend trade-offs, and reason about a system at scale. As with the last round, no AI assistance.
Step 6 — Behavioral
45 minutes · conversation with an engineering leader
A conversation built around specific, real examples from your career, including the times things didn’t go to plan. We’re interested in ownership, judgment, how you handle disagreement, and how you’ve grown from feedback.
How to prepare
- Come ready to go deep on one or two systems you’ve owned end to end, including the scale they ran at and a hard problem you had to debug.
- Brush up on core data structures and algorithmic trade-offs, you’ll reason through problems out loud, not just land on an answer.
- Practice narrating a system design from scratch, and be ready to adjust it as the constraints change mid-conversation.