Tech Abstractions
Interview-grade reasoning for AI and ML engineers.
Tech Abstractions puts production-grade course material and interview practice in one place. Read the theory for free, then practise the rounds where it gets tested: scoping, design decisions and follow-ups that climb from mid-level to staff.
Interview prep teaches answers. Interviews test reasoning.
Engineers memorise system designs and still stall when the interviewer changes a constraint. What gets tested is how you scope, decide and defend a call under follow-ups.
Tech Abstractions practises exactly that, on scenarios drawn from production AI and ML systems.
How it works
Read
Work through free chapters written about production systems, not toy examples.
Scope
Ask the interviewer questions, or state your assumptions. Both earn the same credit.
Decide
Make the design calls and handle follow-ups that climb in difficulty from mid-level to staff.
Debrief
See your level on each dimension, the expert answer, and where to practise next.
What makes it different
One interviewer, written by people
Practice runs on question ladders authored in advance. No language model writes the text you read at runtime.
Assumptions count
Stating an assumption earns the same credit as asking, the way a good interviewer scores it.
Built on production cases
Scenarios come from real AI and ML systems: fraud at scale, serving costs, evaluation and rollout.
Where it stands
Live, as of October 2026.
Built
- Course reader with 11 chapters across three courses, a practice bank and Stripe payments.
Now
- A redesign around a new scenario player, lessons and interview rounds.
Next
- Tracks for ML, AI and LLM inference engineers, and an interview simulator.