Skip to content

For hiring teams

A first-round screen that shows its evidence

Candidates take a real round. You read what they said, what their code did against hidden tests, and how they handled an incident.

Not ready to sign up? Talk to us. We can set up a pilot with you.

How it works

Make it. Send it. Read it.

You make an assessment, send links, and read the results in one place.

  1. 01

    Make an assessment

    Pick a round, the level you hire for and a time limit from 15 to 120 minutes.

    Example

  2. 02

    Send invite links

    One link for each candidate. A link works once, and for 14 days unless you change that. Inviting is free.

    Example

  3. 03

    Read the results

    A score, a label for you to review, the time taken, the integrity signals and the evidence behind them.

    Example

Four rounds

Pick the round you hire for

Candidates work in the same rooms our learners use. The interviewer is neutral: no praise, no hints.

  • 01

    Coding

    A problem you pick, or one drawn once for everybody. Hidden tests run on the final code.

  • 02

    Machine coding

    The candidate designs and writes a small system, then handles a new requirement. About 75 minutes.

  • 03

    System design

    A normal design interview on a problem you pick, with an AI interviewer who asks follow-ups.

  • 04

    Incident

    The candidate is on call. They read the alerts and the logs, find the cause and fix it. We suggest 30 minutes.

What you see

A report you can check

Every score has its evidence next to it. This is a sample with made-up data.

Candidate reportSample, not a real candidate

AI signal for a human reviewer

Judged against the SDE-2 bar.

78/ 100

Hire

A signal for a person to review.

8 of 9

The interviewer's hidden tests, run on their final code.

Evidence by area

  • Problem solving3 / 4

    Started with a brute force version, said why it was slow, then moved to a hash map.

    “Checking every pair is n squared. If I store what I have seen, one pass is enough.”
  • Writing working code3 / 4

    Clean code that ran on the first try. Missed one edge case with repeated numbers.

Quotes are word for word from the candidate's own messages. The server checks them first.

Code runs

  • Python, runFailed
  • Python, runRan
  • Python, submittedRan

Integrity signals Signals, not proof

Tab switches
2
Time away
0:40
Pastes
1
Largest paste
212

Never in a report: the hidden test inputs, or the answers they expect.

Sample data. You also get the candidate's final code and the full chat with the interviewer.

What it is not

Honest about limits

A screen is only useful if you know what it can and cannot tell you.

  • Not proctoring

    No webcam, no screen recording and no browser lock. Candidates read a consent text before they start, and it says exactly this.

  • Not an ATS

    It does not track your pipeline or store CVs. It screens one round and gives you the evidence.

  • Not a hiring decision

    The label is a signal for a person to review. Tab switches and paste sizes come from the candidate’s browser, so they show where to look and are not proof.

Privacy and fairness

Candidates know the rules

Candidates are real people with real accounts. We treat their data that way.

  • The candidate gives consent before anything starts.
  • You choose how long data is kept: 30 to 730 days. The default is 180.
  • Our admin screens do not show your candidates’ reports, chats or code.
  • A candidate can ask us to delete their data, and you can delete it too.
  • The candidate sees their own report only if you allow it.

Pricing

Pay for each candidate who takes the round

Inviting is free. You buy a pack of credits. A credit is used only when a candidate starts and gives 2 real answers, or gets a report.

  • Hiring Pack 10

    10 candidates

    about ₹499 each

    ₹4,990

  • Hiring Pack 50

    50 candidates

    about ₹400 each

    ₹19,990

  • Hiring Pack 200

    200 candidates

    about ₹300 each

    ₹59,990

  • No GST is added: the price you see is the price you pay.
  • Credits do not expire.
  • If a round does not count, the credit comes back.
  • Paying online is not open yet. For now, tell us how many candidates and we add credits for a pilot.

Questions

Before you send a link

Can candidates cheat with AI?

We cannot stop it, and we will not say we can. The report shows signals next to the evidence. Read the evidence, then ask the candidate about it in your next round.

Who can see the results?

The people you add to your team: owners and recruiters (up to 25). Our admin screens do not show your candidates’ reports, chats or code. A person who is not on your team cannot open your page at all.

What does the candidate see?

First a consent text, before anything starts. Then a normal room with a clock and no hints. After the round they see their own report only if you allow it. By default they do not.

What if a round does not count?

You use a credit only when the candidate starts and gives at least 2 real answers, or gets a report. If the candidate leaves too early, or the code runner is down, or we cannot write the report, the credit comes back and you can send the link again.

Do you give an invoice?

You get an invoice for each pack you buy. The exact kind depends on our tax registration, which is not final yet. Write to us if your finance team needs something specific.

Try it with a pilot

Make an account, or write to us. We will help you send the first links.