Day One · An assessment solution by Mercajob

Hire with Mercajob
Join the pilot

AI work simulation

See how they work, before their first day.

Day One puts the candidate in a real working day for 25 minutes: emails, data, documents, a manager, a teammate and an AI assistant at hand. You see what an interview doesn’t show: how they organise, check, decide and use AI.

  • 50 case studies across 10 sectors
  • 25 minutes, in French or English
  • Scored by fixed rules: AI never grades
Join the pilot programme See how it works

Why Day One

Your future hires will work with AI. Assess them with it.

Banning AI during a test tells you nothing about how someone will work on the job. Day One makes it available, just like at the office, and observes what makes the difference: knowing when to use it, what to ask, and not taking its word for it.

  • A real situation, not a questionnaire

    The candidate produces real deliverables: a reply to a client, a calculation, a selection, a recommendation.

  • Realistic traps

    An outdated document, a wrong figure in a summary, a mistake slipped in by the assistant, an email with a hidden instruction.

  • Comparable results

    Same structure and criteria in every sector: the candidate picks a context they are comfortable with, you compare on a common basis.

How it works

25 minutes, like a real morning at work.

The candidate picks a sector, then one of its 5 case studies, and receives a brief from their manager. On screen: an inbox, a data table, documents, the AI assistant and a simulated teammate working on another part of the case.

The 6 tasks of a session

  1. Client replyAn email to handle, sent to the right person
  2. Key figureA calculation from the data
  3. SelectionThe cases to handle first
  4. AnalysisA justified action plan
  5. CurveballThe manager changes the brief: adapt the plan
  6. Wrap-upThe update for the meeting, then a session review
  1. The AI assistant

    A tool, with a budget

    The candidate uses it freely, within a credit limit, as a resource to spend wisely. Once per session, the assistant makes a deliberate mistake.

  2. The teammate

    A colleague who needs them

    They ask for a figure, share an assumption and expect an answer on time. You see how the candidate works in a team.

  3. The data

    Fictional, but credible

    Companies, people and figures are made up. No real data is handled, by the candidate or by the AI.

What is observed

7 criteria, each with its reasons.

Each criterion is computed from the deliverables and the session log, using rules written in advance. You see the suggested score and, above all, why.

  • 20 pts

    Task execution

    The right recipient, the right calculation, the right selection.

  • 20 pts

    Use of AI

    The right moments, precise requests, answers reworked rather than copied.

  • 15 pts

    Critical thinking

    Spotting the wrong figure, the outdated document, the assistant’s mistake and the hidden instruction; protecting personal data.

  • 15 pts

    Analysis and proposals

    Causes identified and actions prioritised, with a clear decision.

  • 10 pts

    Adaptability

    The new constraint is genuinely taken into account.

  • 10 pts

    Teamwork

    The teammate gets the right figure, on time.

  • 10 pts

    Wrap-up

    A clear summary, and an honest review of the session.

Case studies

50 case studies, across 10 sectors.

The sector is a backdrop: it puts the candidate at ease without becoming the challenge. The skills observed are the same everywhere.

  • Banking, insurance

    • Rising customer complaints
    • Launching home insurance for young professionals
    • Bringing customer files into compliance
    • Risky mortgage applications
    • Backlogged car insurance claims

  • Tech, software

    • A wave of customer cancellations
    • Overwhelmed customer support
    • Free trials that don’t convert
    • Recurring production incidents
    • Roadmap and at-risk customers

  • Retail, e-commerce

    • Product returns soaring
    • In-store stock-outs
    • Promo codes: costly and misapplied
    • Falling customer ratings
    • Late furniture deliveries

  • Industry

    • Defective parts on a production line
    • Overdue machine maintenance
    • A key supplier in trouble
    • Workplace accidents at the plant
    • Off-contract purchasing out of control

  • Healthcare, pharma

    • Missed medical appointments
    • Medicines nearing expiry
    • Clinical trial behind on enrolment
    • Long waits for online consultations
    • Late adverse-event reports

  • Energy, environment

    • Carbon footprint: transport underestimated
    • Complaints about electricity bills
    • Wind turbines less available than planned
    • Customers falling behind on waste sorting
    • Stalled renovation grant applications

  • Transport, logistics

    • Delivery rounds exceeding driving-time limits
    • A fuel-hungry van fleet
    • Picking errors in the warehouse
    • Freight disputes to settle
    • Late regional trains

  • Consulting, business services

    • Under-utilised consultants and a new assignment
    • Unpaid invoices and tight cash
    • Fixed-price projects losing money
    • Unhappy clients at a payroll provider
    • Expense claims outside policy

  • Real estate, construction

    • Unpaid rent in property management
    • Energy-inefficient homes to renovate
    • Housing construction running late
    • New-build programme: shaky sales
    • Lift breakdowns in a residential building

  • Tourism, hospitality

    • A half-empty hotel on weekdays
    • Rising holiday cancellations
    • A corporate offsite to finalise
    • Non-compliant kitchen temperatures
    • School groups at a theme park

Day One can be taken in French or English: all 50 case studies exist in both languages. All companies, people and data in the case studies are fictional.

What you get

A session log, not just a score.

For each candidate, you replay the session as if you had been there, then give your own assessment.

  • The session timeline

    Documents opened, requests made to the AI, exchanges with the teammate, deliverables sent.

  • A suggested score, criterion by criterion

    Every point gained or lost is explained. You can adjust it.

  • Your recruiter assessment

    Your notes are kept with the result, so you can decide as a team.

  • A benchmark, when it makes sense

    The result is benchmarked against other sessions once 30 comparable results exist.

Method

AI is a tool in the test. It never judges.

Day One is designed so that the decision stays human, and so it can be explained simply to a candidate, a leadership team or employee representatives.

  • Scoring by fixed rules

    Expected answers are computed from the case-study data, never by AI, and are never sent to the candidate’s browser.

  • A human decision

    The score is a reasoned suggestion. The recruiter reviews, adjusts and decides.

  • Transparency for the candidate

    They know in advance how the session works and what is observed, and can view their own session log.

Pilot programme

Try Day One on your next hires.

Day One is open to a small number of pilot companies. We prepare the session with you, set up access for your candidates and review the results together. The pilot terms are presented during a first call.

  1. 01

    A first call

    Your roles, your sectors, your expectations. We walk you through a full session.

  2. 02

    Your candidates take Day One

    They are informed in advance and receive personal access. 25 minutes, from a web browser.

  3. 03

    You decide, we measure

    You review the session logs. We gather your feedback to improve the tool.

Join the pilot programme Discover Mercajob for companies

Prefer to talk it through? Book a 30-minute call, or write to contact@mercajob.com : we answer within 2 business days. Day One is also available to talents registered on Mercajob, who can share their result with companies.