AI Engineering Readiness

30 minutes. A map that changes what you do next.

We measure the AI delta that gets you hired. Real-world questions across the five areas where AI Engineering is decided, each measured to high confidence. You leave with an area-by-area readiness profile and the exact courses to close every gap. Free. Private.

Sign in to start — it's free

The assessment

What you’ll actually face.

Five short modules — one per area — each adapting from the basics up to the job-ready bar. Around 12–30 questions in total, about 30 minutes. No trivia. Every question is grounded in a real scenario — the kind of thing you’d face on the job.

Scenario judgment~40%

A real production situation — broken agent loop, bad retrieval results, a safety gap. You pick the most surgical fix under the stated constraints.

Example

Your RAG pipeline is returning chunks that pass similarity thresholds but fail fact-verification. The most targeted fix before reranking is…

Code & trace reading~25%

A short snippet, diff, or agent execution trace with a planted defect. You name the defect or the most precise repair.

Example

chunks = sorted(chunks, key=lambda c: len(c.text), reverse=True)
prompt = build_prompt(chunks[:5])
→ What is the most likely problem?
Numeric & expression~15%

A cost or latency calculation, a missing schema field, a regex or SQL fragment. Far harder to guess than four-option MCQ.

Example

An LLM-as-judge scores 200 responses. It agrees with human raters on 150. What is the agreement rate?

5 adaptive modules~12–30 questions~30 minutesOne 24h pause availableNothing to cram

No clock, no rush. There’s no timer and no minimum — take all the time you want on every question. The ~30 minutes is just a typical pace, not a limit. Accuracy is all that matters here.

The report

What you walk away with.

An immediate, domain-by-domain readiness profile. Not a score out of 100 — a map of where you are and exactly what moves next.

Area levels you can act on

Each of the five areas is placed on the L0–L3 ladder — Getting started, Foundations, Developing, and Job-ready (the bar) — by how high you climb in its module. The dashed line on every bar is the L3 threshold. You're Job-ready overall when all five clear it.

Role-fit summary

A derived signal from your domain profile: Agentic-fit, how you'd do in an Agentic AI Engineer role. It's an explanatory pattern, not a separate verdict — it helps you understand what your profile points toward.

Gap-matched course routing

Every area below the Job-ready bar emits a ranked list of courses — hand-picked from the best providers in the world, matched to your specific gap. No generic recommendations. Each course closes a named area.

L0 Getting started · L1 Foundations · L2 Developing · L3 Job‑ready (the bar)

Sample report

Developing.

Strong where it counts — two areas to close to reach Job-ready.

LLM Fluency82L3
Building with AI78L3
Agent Engineering54L2
Trust & Evals80L3
Ship & Scale58L2
Agentic-fitDeveloping

The courses

Gaps become a plan.

Independent by design: every gap in your report maps directly to the best course in the world for closing it — hand-picked from providers builders actually trust.

DeepLearning.AIAAnthropic AcademyOpenAI AcademyCCourseraHHugging FaceMMaven

Ready to see your map?

It’s free. Results stay private. About 30 minutes — one focused session.

Sign in to start — it's free