Free live workshop · Sat 21 Nov · 60 minutes on Google Meet Save my free seat →
For software engineers interviewing for Staff Engineer, Principal Engineer & Engineering Manager roles

Practise the agentic system-design interview. Live.

A free, live 60-minute workshop with Sunil Mathew — 26 years in engineering, leadership at Google, Amazon and Walmart Labs. Answer one real design question, get it scored by an AI evaluator twice, then hear exactly how a senior interviewer would judge it.

  • 100% free
  • No code, no install
  • Live, not a recording
  • Work on your own answer
Save my free seat → Takes about 40 seconds. Joining link shown instantly.
Sunil Mathew

Hosted by Sunil Mathew
Has hired and assessed senior engineers · Founder, The Living Craft

Saturday, 21 November 2026

10:00 IST · 08:30 GST · 04:30 GMT

STEP 1 OF 2 · WHERE TO SEND YOUR LINK

Please add your name.

Please enter a valid email.

With country code. Only used for your joining link and one reminder.

Please include your country code, e.g. +1, +91, +971.

Free. No payment details. Unsubscribe anytime.

Your host has led engineering at
  • Google
  • Amazon
  • Walmart Labs
  • 26 years in engineering
The senior design round

You know the system. The panel is listening for your judgment.

In a senior agentic design round, listing a model, a tool and a memory store isn't the answer. The interviewer wants to hear what you'd let the agent do on its own, what could go wrong, and why you chose this over the alternatives.

Where answers lose the room

You name the components.

The interviewer hears about a model, a tool and a memory store, but still doesn't know which action you'd allow or what could go wrong.

"We'd use an LLM with tool calling, a vector store for memory, and an orchestrator…"

Where answers lose the room

You leave your judgment implicit.

You may have weighed several alternatives. If you only describe the one you chose, the panel has little evidence of how you decided.

"…so that's the architecture."
— and the trade-offs stay in your head.

"What would you let an AI agent do on its own? And what evidence would make you change your mind?"The follow-up question this workshop trains you to answer.

What you'll learn

Sunil's seven-stage template for any agentic design question.

A structure that makes your reasoning visible — so the panel hears decisions, not a parts list.

  1. Frame the decision
  2. Set the agent's boundary
  3. Establish the evidence
  4. Place the controls
  5. Trace what can go wrong
  6. Weigh the alternatives
  7. Say what would change your mind
Learn it live — free →

Then use it on a real question, in the same hour.

How the hour runs

Not a lecture. You do the round.

You develop your own answer before you see Sunil's — the closest thing to the real interview, without the stakes.

Save my free seat →
  1. Opening

    The design problem

    Hear the problem and the level of reasoning the senior round asks for.

  2. Template

    Seven stages, with Sunil

    Walk through the structure senior panels listen for.

  3. Question

    Read the prompt

    Identify the decision you'll need to defend.

  4. Your turn

    Prepare your own answer

    Before you see anyone else's.

  5. Score it ×2

    Run the evaluator — twice

    Compare both runs on the same answer and see what moved.

  6. Sunil's review

    Trace a strong answer

    Which decision mattered, which trade-off it exposed, and the follow-up he'd ask.

  7. Close

    Pick what to practise next

    Leave with one part of your reasoning to sharpen — and ask your questions.

Why a human in the loop

An AI evaluator checks coverage. A senior interviewer checks judgment.

You'll get both — and see where they disagree.

The AI evaluator

Coverage

Points out the topics your answer addressed or missed.

  • Run it twice on the same answer
  • Spot a judgment that moved even though your answer didn't
  • Learn what automated scoring can and can't tell you

Sunil Mathew

Judgment

Explains which decision matters, what trade-off it exposes and the follow-up question he'd ask.

  • Hear the reasoning behind interviewer follow-ups
  • See what separates a strong senior answer
  • Know which part of yours needs more evidence
What you leave with

Six things for your next practice round.

Get all six — free →
  • 01

    A seven-stage structure

    For approaching any agentic design question.

  • 02

    Your own worked answer

    To a real session question — not a toy example.

  • 03

    A first evaluation

    Of where your answer covers the ground and where it doesn't.

  • 04

    A second evaluation

    To compare with the first, and calibrate how much to trust AI scoring.

  • 05

    Sunil's strong answer

    And his explanation of what makes it strong.

  • 06

    One thing to practise

    A specific part of your reasoning to work on next.

Who it's for

Built for engineers going for the senior round.

This is for you if…

  • You're a software engineer interviewing for Staff Engineer, Principal Engineer, Architect or Engineering Manager roles
  • You can already discuss how a system is designed
  • Agentic AI design questions are showing up in your loops
  • You want to hear how an interviewer actually assesses an answer

Probably not for you if…

  • You're looking for a coding tutorial
  • You want a general introduction to AI
  • You want a job guarantee — this is practice, and hiring decisions belong to the employer
Sunil Mathew
Sunil MathewFounder, The Living Craft
Your host

Learn from someone who has sat on the other side of the table.

Sunil Mathew has 26 years of engineering experience, including leadership roles at Google, Amazon and Walmart Labs. He has hired and assessed senior engineers. In this workshop you hear the reasoning behind an interviewer's follow-up questions — the part candidates rarely get to see.

26years in engineering
3leadership roles: Google, Amazon, Walmart Labs
60focused minutes, live
Which decision matters, what trade-off it exposes, and the follow-up question a senior interviewer would ask about it.
Questions

Before you join

Anything else? Email founder@thelivingcraft.ai.

Is it really free?

Yes. The workshop is free. If you later want personal feedback for a specific interview round, the Focused Mock Interview is a separate paid service.

What time is it where I am?

Saturday 21 November 2026, 10:00–11:00 IST (08:30 GST · 04:30 GMT). In the US that's Friday evening, 20 November.

What do I need?

A browser and ChatGPT, Gemini or Claude open in another tab — you'll paste your answer there during the session. A one-page guide arrives the day before and the evaluator link an hour before.

Do I need to code?

No. You'll work on explaining a system-design decision. Use the workshop example and keep confidential employer or interview material out of the AI tools.

How do I join?

On Google Meet. You'll see the joining link as soon as you register, and we'll send it again with a reminder on WhatsApp.

Will this get me the job?

No one can promise that, and we won't. This is interview practice — hiring decisions belong to the employer. What you get is a clearer structure and a sharper sense of how your answer lands.

Free · Live · 60 minutes

Rehearse the round before it counts.

One real question. Two AI evaluations. One senior interviewer's honest read. Saturday, 21 November.

Save my free seat →

Sat 21 Nov · 10:00 IST · 08:30 GST · Google Meet

Free live workshopSat 21 Nov · 60 min
Save my seat