ChannelPulse
AI interviews
The newest rounds in the loop, for engineers, product managers, forward-deployed engineers and engineering managers. Some companies now expect you to work with an AI assistant during the interview and grade how you do it; others put an AI on the other side of the call as the first screen. This course covers both — who runs them, how they are scored, and a prep path for each role.
The AI-assisted round isn't graded on whether you used the tool or avoided it. It's graded on whether you stayed in charge of it — your plan before the first prompt, small requests you can check, and every piece read, run and tested out loud. This course drills that loop across code, design, product cases and the AI-run screens.
The new rounds — what changed and how the AI round works
Two different things get called an "AI interview": rounds where you work with an AI assistant, and screens where an AI interviews you. This module separates them, says which companies run which (confirmed versus reported), and walks through the reference format — the three-phase AI-assisted coding round.
-
1
What changed — AI-assisted rounds and AI-run screens
Two unrelated things now share the name "AI interview". One is a round where the company hands you an AI assistant and grades how you use it; the other is a screen where an AI is the interviewer. This lesson separates them, shows who runs each and how sure we are, and sets up the rest of the course. Free -
2
The AI-assisted coding round, phase by phase
The structured AI round is sixty minutes in a multi-file repo with an AI chat beside the editor, run as three phases on one problem. This lesson walks the format minute by minute, says what each phase is grading, and shows how much of it you actually need to finish. Free
Engineers — the AI round, debugging, design and open builds
How engineers are graded when the AI is in the room: the plan, prompt, verify, narrate loop on a real repo, finding bugs and writing tests in code you didn't write, and the design interviews and multi-hour builds that AI-native companies now run.
-
1
Working with the AI — plan, prompt small, verify, narrate
Every source that describes how these rounds are graded lands on the same loop — your plan before the first prompt, small requests you can review in one read, a run after every change, and your reasoning said out loud. This lesson takes one reported problem and plays it twice, the way that fails and the way that passes. 17 min · app -
2
Debugging and testing code you didn't write
The AI rounds put you in someone else's repo far more often than classic rounds did — a bug-fix phase, a repo of failing tests in an online assessment, a pull request to review, a module to explain. This lesson gives you a five-minute orientation, a triage method for failing tests, and a way to review AI-written code without trusting the AI's summary of it. 16 min · app -
3
Design rounds, open-ended builds and the long build
Not every AI round hands you a repo and three phases. Some give you an empty repository and keep adding requirements; some give you hours to build a product with any tool and then make you defend it; and Meta now runs design interviews with the assistant in the editor. This lesson covers all three and the habit they share — the structure is yours, the typing is the tool's. 15 min · app
Product managers — product sense with AI, LLM features and prototypes
The PM versions: a product-sense case with an AI assistant on screen, designing and measuring a feature built on a model, and the prototyping rounds where you build with an AI tool in the interview.
-
1
Product sense with an AI on screen — and product sense about AI
Two new PM rounds share a name. In one, you run an ordinary product-sense case with an AI assistant open and your screen shared, and you're graded on your judgment about its output. In the other, there's no tool at all — the product is an AI product, and the case adds model limits, cost and safety to everything you already do. This lesson covers both. 16 min · app -
2
Designing an LLM feature, and proving it works
The AI PM technical round asks you to design something a model powers — and then to say how you'd know it's good. This lesson gives you a design checklist built around failure modes, a way to choose between rules, classic ML and an LLM, and an eval plan you can describe in two minutes, plus the triage tree for a model-backed feature that's underperforming. 17 min · app -
3
Prototyping rounds — building with an AI tool as a PM
Some PM loops now include a round, or a take-home, where you build a working prototype with an AI tool such as v0, Lovable, Cursor or Replit. The evidence is thinner than for any other round in this course, so this lesson says plainly what's known, then gives you a structure that holds up whatever the exact format turns out to be. 12 min · app
Forward-deployed engineers and engineering managers
The two roles where the AI is mostly the subject of the interview rather than a tool in it: building and fixing LLM systems in front of a customer, and leading a team through AI adoption.
-
1
Forward-deployed engineers — customers, decomposition and LLM systems
Forward-deployed engineer loops look like software engineering loops with two extra rounds that filter out most candidates — decomposing a vague customer goal, and handling a customer in a role-play — and for AI companies, a technical deep dive on building LLM systems that work. This lesson covers each round and the numbers you should be able to quote from your own work. 18 min · app -
2
Engineering managers — the AI coding round and leading AI adoption
Managers meet the AI interview twice. Some loops now give them the same AI-assisted coding round as engineers, sometimes as the only coding round; and leadership rounds increasingly ask how you've rolled AI tools out to a team, measured the result and handled what went wrong. This lesson covers both, and is explicit that the second is the least documented part of the course. 15 min · app
AI-run screens and your prep path
The screens an AI conducts before you meet a person — how they work, what they score and how to do well — and a two-week prep path for each of the four roles.
-
1
AI-run screens — what they score and how to do well
For many applicants the first interview is now with an AI — a voice or video agent that asks role questions, follows up, and writes a transcript and a score for a person to review. Candidates mostly dislike them, and some walk away. If you take one, this lesson explains what it actually measures and the handful of habits that change the outcome. 11 min · app -
2
A two-week prep path for each role
Fourteen days, laid out per role — engineer, product manager, forward-deployed engineer, engineering manager — with what to drill each day, what to measure, and the recruiter questions to settle before you start. It pulls the previous eleven lessons into one plan, so you can start tomorrow. 12 min · app
Work through it with feedback
Reading the pattern is step one. The app runs you through it: editable code cells, the question bank, and a mock loop that grades your answer.
Log in to open the curriculum