How our free AI career consultation works

1. The questions we ask

About you

• Which best describes you right now?
• How many years of work experience do you have?
• What is (or was) your main area of work?
• Which AI role are you in today?
• What is pushing you towards AI right now?

Your skills today

• Which programming languages can you write code in?
• How would you rate your Python?
• How comfortable are you with SQL and data?
• How much cloud / deployment experience do you have?
• What have you done with AI / ML so far?
• How comfortable are you with math (statistics, algebra)?

Quick reasoning check

• The average of four numbers is 10. Three of them are 8, 9 and 11. What is the fourth?
• A model is 99% accurate at spotting a disease that only 1% of people have. Is it a good model?
• An AI chatbot gave correct answers last week. This week, with no code changes, its answers are wrong. What do you check first?
• A retail client says: "Use AI to stop customers leaving us." What is the best first step?

How you like to work

• Which kind of workday sounds most like you?
• Which parts of AI excite you?
• Which industry do you know best?

Goals & time

• What do you want to achieve?
• How many hours a week can you realistically study?
• By when do you want to be job-ready?

Some questions only appear when relevant. For example, students aren't asked about their current job.

2. How we validate your answers

• Reasoning check: four short questions (numeric reasoning, data intuition, systems thinking, problem framing) test the abilities people tend to over- or under-rate.
• Maths calibration: rating your maths as strong but missing the maths question lowers it to "okay". Rating it as weak but solving the question raises it slightly.
• Consistency rules: for example, listing Python as a language but rating it "none" is treated as basic Python. Claiming model training without any programming language is treated as having used AI tools.
• Every adjustment is shown to you on the results page, with the reason.

3. How each AI role is scored

Each role has a set of weighted signals. A role's fit percentage is its score compared with a near-ideal profile for that role. The strongest signals for each role are below (+ raises the score, − lowers it). These are the same sentences you see as reasons on your results.

• AI Foundations Path
  • +6 You're early in coding, so a strong foundation is the fastest route
  • +5 As a student / graduate, a structured foundation path gets you job-ready
  • +3 Your goal is a first job in tech — this path targets junior AI roles
  • +2 Little work experience yet — the foundations matter most
  • -7 You already have engineering experience — a beginner path would waste it
  • -6 You already train models — you're past the foundations
• ML Engineer
  • +4 You've already trained ML models
  • +4 You're excited by training and tuning models
  • +3 Your Python is at a working level
  • +3 Your maths comfort supports model work
  • -4 ML Engineering needs solid Python, which you're starting from zero
  • -3 ML Engineering needs comfortable statistics
• Data Engineer (AI focus)
  • +7 You already work with data pipelines / analytics — a direct bridge
  • +5 You enjoy working with data
  • +4 Your SQL is strong
  • +3 You like building reliable systems
  • -4 Data engineering is SQL-heavy and you're new to it
• AI Product Engineer
  • +6 You already build apps — AI features are the next layer
  • +5 You want to add AI features to apps
  • +5 You love building things people use
  • +3 JavaScript / TypeScript is the main stack for AI product UIs
  • -4 This is a coding role and you don't code yet
• MLOps / AI Platform Engineer
  • +8 Your DevOps / operations background transfers almost directly
  • +5 You already have cloud and deployment skills
  • +5 You're interested in AI infrastructure
  • +5 You like making systems reliable at scale
  • -4 MLOps is infrastructure-heavy and you haven't deployed to cloud yet
• LLM / Agent Engineer
  • +6 Autonomous agents excite you
  • +5 You want to automate real work with AI
  • +4 You've already built with LLM APIs
  • +4 Backend / automation experience is exactly what makes agents reliable
  • -5 Agent engineering is a coding role and you don't code yet
• AI Research Engineer
  • +6 You already have hands-on model training experience
  • +5 Your maths is strong
  • +5 You're excited by how models learn
  • +5 You love deep technical problems
  • -6 Research roles need strong maths, which you rated as weak
  • -4 Research engineering needs advanced Python
• AI Forward Deployed Engineer
  • +7 You enjoy solving customers' problems hands-on
  • +5 You care about business impact
  • +4 Your industry knowledge is a big advantage with clients
  • +4 Your years of experience build client trust
  • -5 Forward deployed roles need hands-on coding
  • -2 Client-facing roles usually want some work experience first
• AI Product Manager
  • +8 You already manage products, projects or requirements
  • +6 You like deciding what to build and why
  • +4 Domain experts from outside IT make strong AI PMs
  • +3 Your industry knowledge helps you find the right AI use cases
  • -3 PM roles usually want some work experience first

4. Readiness, skill gaps and timeline

Each role lists target levels for Python, SQL, cloud, hands-on AI and maths. Readiness is how much of that you already have, and anything below target becomes a skill gap with advice. Roadmap steps that cover skills you already have are marked fast-track and shortened to a quick revision. The remaining study hours are spread over your weekly hours to give real dates, which are checked against your target date.

5. Limits of the recommendation

• It relies on your self-ratings. The reasoning check reduces this problem but doesn't remove it.
• Salary ranges are indicative, drawn from public sources (Levels.fyi, Glassdoor, AmbitionBox and 2026 salary guides) that often disagree.
• It is guidance to focus your learning, not a job guarantee. Use the top 2–3 matches as a shortlist and validate with a small project, as described in how to choose the right AI role.

Frequently asked questions