The Rounds and What They Measure
A .NET + AI interview loop is not one exam — it's a series of rounds, each scoring something different. Candidates fail when they give a system-design answer in a screen, or trivia in a design round. Know what each round is for, and calibrate.
The typical loop
The recruiter/HR screen (20–30 min) measures exactly two things: can you describe your work clearly to a non-engineer, and do your logistics fit (start date, location, clearance/background check — common in public-sector roles). It is not technical. Red flag here: rambling a 10-minute architecture story at someone who wanted "senior ASP.NET Core developer, 6 years, EF Core and Azure DevOps."
The technical screen (45–60 min) is usually one engineer probing depth: "Walk me through
what happens when a request hits your controller." "What's the difference between IEnumerable
and IQueryable?" They're scoring fluency — do you talk about your daily stack the way
someone who actually uses it does. Hesitating on your own tools is the fastest fail.
The system-design round (60 min) hands you something open-ended: "Design a service that ingests thousands of transcripts and flags them for reviewer attention." Nobody expects a perfect design. They score structure: do you clarify requirements, name components, discuss trade-offs, and handle a curveball ("now the AI service goes down mid-batch — what happens?") without freezing.
The behavioral round measures whether they want you on the team: conflict, failure, ownership. Public-sector panels often run this as a structured scorecard — same questions to every candidate, notes graded later — so specific, complete stories beat charm.
Who is across the table
Match depth to audience. The engineering manager wants judgment and delivery; the senior engineer wants depth and will chase your first vague statement; the product or program person wants you to explain AI-assisted review to a non-technical stakeholder without jargon. Ask who you're meeting beforehand — it's a normal question and it changes your prep.
The meta-signal
Every round secretly measures the same thing: what is it like to work with you when the problem is unclear? Interviewers simulate that by leaving questions underspecified on purpose. The candidate who asks one clarifying question before answering almost always outscores the one who launches into a memorized monologue.
Practice prompts — answer these out loud:
- In two minutes, describe your current role to a recruiter who doesn't code.
- You're told tomorrow's panel is "a lead engineer and a business analyst." What do you prepare differently for each?
- An interviewer asks a question you think is ambiguous. Say the exact sentence you'd use to clarify it without stalling.