One capability profile. Targeted career matching.
Ukandu turns a 29-capability profile into a ranked list of roles that genuinely fit you, with a clear explanation for every match. No black box, no personality quiz, no “people like you also liked”. Just a deterministic engine that compares what you can do against what each role actually needs, and explains every result in plain English.
Four stages. One result.
Your capability profile is the output of a structured pipeline. Two stages capture what interests you, one measures what you can do, and one matches you against real workplace roles.
No black box. Just clear reasoning.
The matching engine does not “decide” anything. It compares. Your capability profile meets each role's capability profile, and the engine ranks the roles where what you bring most lines up with what the role needs most. Every score traces back to specific, named capabilities.
The same answer. Every run.
Two profiles with identical capability levels get the same ranking. Two roles with the same capability priorities get the same fit score. No drift, no model retraining, no “the algorithm has changed”. That is what deterministic means, and it is the foundation everything else sits on.
Your strengths. The role's priorities.
Every role has a small set of capabilities that genuinely make or break the work, and a larger set that contributes around the edges. The engine looks at where the role's priorities and your strengths meet, where there are gaps on the things that matter most, and how cleanly the shapes line up overall.
The further you sit below what a role really needs on a critical capability, the more it counts against the fit. A small gap on something peripheral barely registers. That is what makes the ranking shape-sensitive, not just a tally of strengths.
Every score, traced to a capability.
Because the engine is deterministic and operates on named capabilities, every fit number unpacks into specific reasons. On the results page, each role comes with several layers of explanation, all generated from the same profile that produced the score.
- i.Why it suits you. the capabilities where what the role values most and what you bring most genuinely line up. The shape of the overlap.
- ii.Watch out for. capabilities the role leans on heavily where you sit below what it really needs. Honest gaps, surfaced up front.
- iii.A day in the life. what the role actually looks like in practice, anchored to the same capability set so you can map it back to your profile.
- iv.Does this fit you. a per-role rating widget so your gut reaction calibrates the engine over time.
The lines we don't cross.
We have made some deliberate calls about what Ukandu will and will not do. They are what makes the result trustworthy at scale, by you, by recruiters, by enterprise, eventually by regulators.
- Deterministic. Same input always produces the same output. No drift, no opaque model updates.
- Capability-based. Built on 29 measurable workplace capabilities, not traits or vibes.
- Priority-aware. Roles are not matched on keyword overlap. They are matched on the capabilities that genuinely make or break the work.
- Explainable. Every score traces back to specific, named capabilities. You can read the reasoning behind every match.
- Auditable. Methodology is published. Role weights are versioned. Results survive scrutiny.
- Not a personality test. We do not tell you which animal or colour you are.
- Not a black-box recommender. No "trust us, the model said so".
- Not a collaborative filter. We never recommend a role because "people like you also liked it".
- Not a resume parser. Your past job titles do not define what you are capable of next.
- Not a job board. We surface fit. You choose what to do with it.
Map your capabilities. Find the roles.
Around 30 minutes end to end. Stop and resume any time. Your result lands the moment the engine is done.