A three-hour career workshop

Prepare for the interview, not just the syllabus.

ML interviews test more than algorithm definitions. A candidate must identify the right roles, explain concepts clearly, defend project decisions, design complete systems, solve coding problems, and communicate credible evidence.

Why this session is personal

I used versions of this preparation system while preparing for and clearing interview processes at Qualcomm, HARMAN, Amazon and Walmart. The process changed between roles, but the demanding parts stayed remarkably similar.

What actually made the difference: not one clever answer or a secret question list, but repeated revision, explaining concepts aloud, defending project decisions, practising coding, sitting through uncomfortable mocks and correcting the weaknesses they exposed.

This experience does not guarantee another person the same outcome. It does provide a tested preparation process that can be adapted, measured and improved.

Opening question

If an interview were scheduled for tomorrow, which stage would expose your biggest gap: resume screening, ML concepts, project discussion, coding, system design, or behaviour?

The complete interview journey

Choose the role
Build evidence
Find opportunities
Prepare deeply
Practise aloud
Apply thoughtfully
Clear screening
Solve interviews
Review feedback
Improve the weakest stage

Central idea: preparation is a system. More study does not automatically improve results if the real bottleneck is weak evidence, unclear answers, unsuitable roles, or insufficient practice.

Three-hour teaching plan

00–10
Choose the role
Role families and interview loops
Output: one target role
10–35
Revise ML effectively
Topic map, retrieval and revision cycles
Output: revision matrix
35–55
Answer ML questions
Structured answers and follow-ups
Output: one spoken answer
55–80
Design an ML system
Business goal through monitoring
Output: pipeline sketch
80–90
Break
Collect questions
90–115
LinkedIn and resume
Positioning, evidence and defensibility
Output: improved bullet
115–140
Find and decode jobs
Search strategy, JD analysis and tailoring
Output: JD evidence map
140–153
Referrals and outreach
Output: one message
153–173
Mock interviews and STAR
Output: one mock answer
173–180
Coding plan and commitment
Output: next three actions

What students should bring

One real job description

A role they would genuinely consider applying for.

Current resume

Even an incomplete resume gives the workshop something concrete to improve.

One project story

A project they can explain through problem, decisions, evidence and learning.

The pages

1

Target the interview

Roles, expectations and interview loops.

 2

Revise ML

A role-weighted topic map and repeatable study cycle.

 3

Answer clearly

Layered answers, examples and follow-ups.

 4

Design a system

One complete pipeline from objective to retraining.

 5

Present evidence

LinkedIn, resume bullets and project defence.

 6

Find aligned jobs

Search, decode, decide and tailor honestly.

 7

Reach people

Referrals, cold outreach and follow-ups.

 8

Practise interviews

Voice mocks, difficulty control, STAR and feedback.

 9

Prepare coding

Patterns, communication and a four-week plan.
Community discussion

Questions, experiences and interview preparation

Ask a question about the material, share what worked in your interview preparation, or suggest an area that deserves a deeper lesson. Comments are public and require a GitHub account.

Begin: choose the target