Product · Skill assessments
Verified, not self-declared
Five questions written for one specific level, four of them correct to pass, on a timer. A verified skill on a profile means someone answered for it — which is the only reason anyone else should trust it.
- Written for the rung being tested
- Every question kept
- The timer is part of it
Advanced verification
Advanced verificationReady for today’s challenge?
Solution Architecture
Advanced
5
Questions
30 sec
Per question
4/5
To verify
Written for the rung being tested
An Expert question is not a Beginner question with longer words. Each level asks a different KIND of question, so passing at Advanced says something Beginner cannot.
Every question kept
The question bank holds everything ever asked, with its options, its correct answers and the reasoning for each one. Admins can read them, retire the weak ones, and import their own.
The timer is part of it
Seconds per question is a setting, and the same clock runs in the preview on this site as in the real thing. Time pressure is what stops an assessment being an open-book exercise.
What a verified skill actually means
A verified skill is one where somebody answered five questions written for that specific level and got at least four of them right. That is the whole claim, and it is deliberately narrow. It does not mean a manager agreed. It does not mean the person has done the job for three years. It does not even mean they are good at it.
It means they could answer, under a clock, without notes. Everything else on a profile is context. This is the part you can check, and the reason anyone else has grounds to trust it.
This is the skills verification layer of the platform: the mechanism that turns a claim on a profile into something demonstrated, and the reason a level here means the same thing in one company as in another.
Why five questions, and not fifty
Long tests do not get finished. Five questions with a four-of-five pass mark is short enough that people take it voluntarily and strict enough that guessing does not survive it: one wrong answer is recoverable, two are not.
The bar is identical for everyone, at every level, in every company using the product. That is what makes one person's Advanced comparable to another's, which is the only thing that makes a skill level worth recording at all.
Question 2 of 5
0:21AKeep the codebase as it is
BA modular monolith
CSplit into microservices
Tested at the level you claim
An assessment tests the level you claim. Claim Advanced and Advanced is what you are asked, with no need to sit Beginner and Intermediate first. Once a level is verified, the next assessment is the rung above it. Verify Expert and there is nothing further to prove.
A claim nobody has tested stays a claim until an assessment verifies it, and a verified level means exactly one thing: that person answered that level's questions, four of five, under the clock. It is evidence of that level, not a record of every rung below it.
Next to verify
Solution Architecture
Change Adaptability
Change AdaptabilityCapacity Management
Capacity ManagementThe points follow the same logic
Levels are worth 25, 50, 100 and 200 XP by default, and administrators can edit that ladder. Climbing pays only the difference, so moving from Intermediate to Advanced pays what Advanced is worth minus what was already banked.
Re-verifying a level you already hold pays nothing. You cannot farm one skill for points, which is what keeps a leaderboard about progress rather than repetition, and keeps XP from quietly becoming a second, worse measure of capability.
4 / 5
Advanced verified
+50 XPThe question bank is yours
Every question ever asked is kept, with its options, its correct answers and the reasoning behind each one. Administrators can read the bank, retire questions that turn out to be weak or ambiguous, and import their own.
Nothing here is a black box you are asked to trust. If a question is unfair, you can find it, read why it was marked the way it was, and remove it. A test nobody is allowed to inspect is not evidence, it is an assertion.
Correct answer
A modular monolith with clear boundaries
Your answer
Split into microservices
What the timer is for
Seconds per question is a setting, defaulting to thirty. The pressure is not there to catch people out. It is there to stop an assessment quietly becoming an open-book exercise, where the thing being measured is search speed rather than capability.
The same clock runs in the sample question on the home page as in the real thing, which is deliberate: the demo is not a gentler version of the product.
Where AI is, and is not
AI drafts questions. It does not decide who passed. Generation is scoped to that one job, every call is logged and auditable, and an administrator sees the shape of a bank before it reaches an employee.
Scoring is four of five, applied identically every time, by code you can read. The judgement stays mechanical on purpose. An assessment whose pass mark could shift depending on who or what was reviewing it is not evidence of anything.