How it works

What is Simverse, and how does the evaluation work?

Simverse is a platform of realistic, hands-on job simulations for technical roles. You step into a working day at a fictional company, do the real work — writing and testing code, reviewing pull requests, handling incidents, answering colleagues — and get a report on the skills you used. This page explains, transparently, how the simulations and scoring work, and what they can and can't tell you.

What a simulation is

Each simulation is a working day at a fictional company. You get a real workspace — an inbox, team chat, tickets, data and a code editor — laid out like the tools most teams use. Your manager has already messaged you. You read the brief, do the work, and make decisions. People react, and deadlines move. It isn't a quiz or a multiple-choice test; it's the job, compressed into about 20–60 minutes.

What we measure

We look at what you actually did, not just whether a final answer was right. Depending on the role, that includes:

• The code you wrote, and whether your tests caught the edge cases a shallow fix would miss.
• How you handled a risky pull request — what you approved, flagged or blocked.
• How you debugged from a symptom back to the cause.
• The decisions you made when the task was ambiguous, and the trade-offs behind them.
• How clearly you communicated with the people in the simulation.

Every point in your report is tied to evidence from your own work in that simulation — the specific moments behind it — so you can see why you scored the way you did.

How scoring works, and where AI fits in

Wherever the work has a clear right answer — does the code pass the hidden tests, does the query return the correct value — we check it deterministically by running it. For open-ended work, like a written update or a judgement call, we use AI to interpret what you wrote against the things the role cares about. The AI reads your work to score it; it does not invent facts about you. We combine these into a readiness or skill signal per role.

What the scores mean — and their limits

Your scores are a practice signal: they show where you did well and where to improve, so you know what to work on next. They are honest about uncertainty, and they have real limits:

• They are not a prediction of whether you'll be hired. Hiring depends on the role, the team and the interview on the day.
• They are not shared with employers. Your results are yours.
• A simulation is a realistic practice environment, not a specific company's codebase, team or pressure.
• One run is a snapshot. Doing more simulations, across more situations, gives a truer picture than any single score.

Fictional by design

Every company, person and dataset inside a simulation is fictional. The tools are laid out like the ones real teams use, but Simverse isn't affiliated with any of those companies. This lets you practise realistic work safely, without touching any real employer's systems or data.

Privacy and your data

We keep your account, résumé (if you upload one), simulation history and results so the product works for you, and you can delete all of it yourself at any time. Full details — what we collect, why, how long we keep it, and how to request deletion — are on our privacy page.