I rolled out an ERP system across fourteen branches of the company.
A team of 25, operating costs down 18%.
PRINCE2 Practitioner, certified in 2019.
Forty-seven applications.Forty-seven interviews.
CV screening narrows the pool before anyone asks a first question. Hi-Po runs a structured voice interview with every candidate, turns it into a competency profile across four areas and sends every one of them a substantive answer. The rejected ones too.
- Interview length
- 30–60 min
- Languages
- over 100
- Assessment areas
- 4K
One opening, two processes
Illustrative diagram. Demo data.
A problem I have
watched for years.
A recruiter cannot meet dozens of candidates in a single hire. So the shortlist gets made from CVs — and a CV says less and less about what a candidate is actually worth. More and more people use AI to get cleanly “past” the ATS.
A good CV correlates less and less with a good candidate. The people who refuse to play that game are usually gone before anyone has heard them.
Hi-Po came out of that observation. A platform that runs a structured interview with the candidate and builds a full competency profile across four key areas, mapped to the role, the position and the industry — instead of leaning on a CV and on the narrow, pre-filtered handful of people a recruiter had time to meet.
With Hi-Po you can interview as many candidates as apply and pick the best on the strength of the interview, not the claims in a document. Recruiters get their time back, and candidates get concrete feedback whatever the outcome.
Before the first question is asked,
Hi-Po has read the CV.
CV Intelligence analyzes the document before the interview and drafts verification questions taken straight out of it. You cannot predict them from the job ad, so you cannot rehearse them.
- 1
Which ERP modules did you roll out yourself, and which ones did the vendor's implementation team take over?
- 2
Where exactly did that 18% come from — renegotiated contracts, headcount structure, or a change in the process?
- 3
Which project did you run under that method after 2019, and what was different about it compared with how you worked before?
CV content is checked for attempts to steer the model (prompt injection). A detected attempt goes into the report, not into the prompt.
This is where every other process
narrows to eight.
Hi-Po starts an interview with every application — in parallel, at a time the candidate picks, in the candidate's own language. Throughput stops being the reason somebody never got heard.
0/ 47interviews running
- Interview time
- 30–60 min
- Languages
- over 100
- Method
- STAR + case
- Record
- full transcript
Preview mode — demo data.
And this is what
one of them sounds like.
The assistant runs the interview STAR-style, presses for specifics and comes back to what it read in the CV. Below is one interview, replayed from its transcript.
question from the CV
Demo transcript. The interview itself is conducted by voice.
- Hi-Po
You said the ERP rollout covered fourteen branches. Which of them pushed back hardest, and what did you do about it?
question from the CV - Candidate
The Katowice branch. They had built their own warehouse system in-house and treated it as their property. Instead of switching it off straight away, we ran both in parallel for a quarter.
- Hi-Po
What exactly convinced them to switch the old system off after that quarter?
follow-up - Candidate
The month-end close times, side by side. Theirs took eleven days, the new one took four. I did not win that argument — their own report did.
- Hi-Po
Describe a situation where a rollout decision of yours turned out to be wrong.
Leadership · STAR - Candidate
I scheduled the implementation team's rotation in the middle of a quarter. It cost us two months. Since then I plan staffing changes at the boundary between reporting periods.
Demo transcript. The interview itself is conducted by voice.
Instead of an opinion —
a 4K competency profile.
Every answer lands in one of four areas mapped to the role. The recruiter sets how much each area weighs for a given position; AI proposes the questions, a human approves them.
- Business focusweight 35%88
Three decisions with a stated impact on margin; the numbers line up with the role description.
- Systems and processesweight 20%74
Solid on inherited processes, weaker on ones designed from scratch; no example of measuring the effect.
- Leadership and managementweight 30%91
Specific conflict situations broken down STAR-style, naming his own mistake.
- Domain expertiseweight 15%82
Confident on operational finance; answered the CV verification question without hesitation.
A ranking that accounts for
every position on it.
Pick a criterion and the list reorders in front of you. Under every row sits the reasoning for the score and the piece of transcript it rests on.
CandidateBusinessProcessesLeadershipExpertise
Top score in the area weighted at 30%. Three management situations broken down STAR-style, each one naming his own mistake and what he took from it.
That time I made the wrong call rotating the team mid-quarter. It cost us two months; since then I plan changes at the boundary between reporting periods.transcript · 00:24:11
Very strong leadership alongside the lowest domain expertise in the pool. Worth considering for a role with solid expert backup.
I took the team over after two key people left. I spent the first six weeks on nothing but one-to-one conversations.transcript · 00:35:20
An even profile with no obvious gap. Answered the CV verification question about his certification precisely, with dates.
I used that method on two projects; on the third we dropped it deliberately, because the documentation overhead outweighed the benefit.transcript · 00:27:35
The strongest business head in the whole pool, but his process answers stay at the level of claims, with no example of an actual rollout.
We raised the line's margin by walking away from two contracts, not by cutting costs. It was an unpopular decision.transcript · 00:18:02
Strongest on process and on substance. Weaker on leadership: his examples are project teams, not line teams.
We cut the month-end close from eleven days to four by moving balance reconciliation to a weekly step.transcript · 00:31:47
An expert profile. His management answers stayed general, with no specific situation named despite two follow-up questions.
Managing a team is above all communication and clear goals.transcript · 00:22:58
Demo ranking. AI recommends, the recruiter decides — the system never hires or rejects anyone on its own.
One narrowing.
A human one.
The axis narrows only here — at the recruiter, who makes the call and has something to defend it with. Going the other way, forty-seven answers come back out.
Every candidate gets substantive feedback: what went well, what was missing and which criteria they were assessed against — whether or not they got an offer. That is a product commitment, not a marketing feature.
Throughput is not the same thing
as decision quality.
| Criterion | CV screening | Hi-Po |
|---|---|---|
| Basis for the assessment | Keywords in a document | Answers to verification questions drawn from the CV |
| Interview reach | A narrow group that survived screening | The whole applicant pool |
| Record of the decision | A recruiter's note | Transcript and the reasoning behind the score |
| Feedback to the candidate | Usually none | Substantive, for everyone |
| Risk control | Depends on who is reading | Built-in safeguards and transcript audits |
Designed with
the AI Act in mind.
The decision belongs to a human
No hiring decision is made automatically. The AI issues a recommendation and shows what it based it on; the recruiter can overrule it without justifying that to the system.
Everything can be checked
The interview transcript and the reasoning behind every point of the score are always available. That is the answer to “on what basis”, ready before anyone asks.
Each organization's data is isolated
Tenant isolation is enforced at the database level, not in application queries. Non-discrimination is a product function, backed by regular transcript audits.
The regulatory status of HR systems is under ongoing legal review as European Commission guidance develops. We do not claim certification or compliance — we claim that we design for these requirements.
Why HRBPs
choose Hi-Po.
Better results than a classic ATS
An ATS filters keywords in a CV. Hi-Po runs a competency interview across four areas, lasting 30 to 60 minutes depending on the number of questions, and identifies real potential rather than a talent for writing CVs.
Every candidate gets substantive feedback
Hi-Po is one of the few platforms on the market that tells every candidate, the rejected ones included, exactly what went well in the interview and what fell short of the expectations for the role. Your employer brand gains even from a “no”.
A full audit trail for the decision
The full interview transcript and the reasoning behind the score are always available — to the recruiter and to the candidate alike. Every rating can be traced back to the statement it came from.
Your High Potentials
are waiting to be interviewed.
We will run the whole process on one of your own job ads — from the CV landing to the ranking with its reasoning. No implementation, no commitment.