Quasar

A hiring pipeline, built and running.

Résumé screening, proctored tests, coding rounds and AI voice interviews, in one product.

Make an offer Watch the 2026 demo
Asking price Open to offers

Source, two apps, deploy configs, handover.

Applicant board for a Backend Engineer role: five candidates with screening, MCQ, coding, tech and HR scores, and their status.
One job’s board. The product set every score.

Applied to hired in five stops.

Each round grades itself and moves the candidate on.

  1. Screening

    AI scores each résumé against the job. Below your threshold, it stops.

    AI screening review for Aarav Mehta: 91% match, matched skills and missing skills.
    91% match, gaps listed.

  2. Multiple choice

    Questions written by AI from your job description, taken under proctoring.

    A candidate's MCQ round: question three of ten, a navigator, and a badge reading two violations.
    Two answers, two violations.

  3. Coding

    Timed code in JavaScript, Java or C++, run against hidden tests.

    Coding round: a Two Sum solution with four of four test cases accepted and a score of 100%.
    Four of four tests pass.

  4. AI interviews

    Gemini Live voice interviews, technical and HR, scored on STAR.

    Technical interview result: score 8.5, STAR scores per dimension and the interview transcript.
    Round one: 8.5 of 10.

  5. Recruiter call

    You meet only who passed, ranked by combined score.

    The candidate's pipeline tracker with every round marked passed.
    Every round passed.

Draw the route once per job.

Rounds
On or off, plus extra technical rounds.
Pass marks
Per round, screening included.
Windows
Each round opens and closes on set dates.
Interviewers
Four AI personas to choose from.
The job posting form's pipeline step: MCQ round, coding round and technical interview rounds with durations and pass marks.

It watches the test, too.

Five critical violations, or zero trust, end the round.

SeverityWhat triggers itTrust
CriticalTab switch, leaving full screen, DevTools−5
WarningCopy, paste, right-click, shortcuts, second monitor−2
StartEveryone starts at full trust100
Proctoring report: trust 96, two violations, a right-click and a keyboard shortcut, each timestamped in the MCQ round.
The test above, as the recruiter sees it.

Candidates practise before they apply.

A free practice side brings candidates to your jobs.

Choose your interviewer: HR manager, FAANG engineer, hostile panel or aggressive startup founder.

Pick a role and an interviewer, then talk.

Progress dashboard: overall score trend across sessions and a STAR radar chart.

STAR scores, trends, XP and 25 badges.

AI coach answering a request for backend roles with a matching job posting and a fit table.

The coach finds open roles and explains the fit.

Résumé against job description: an 85% strong match with a written summary.

Any résumé against any job, with the gaps named.

What changes hands.

Counted from 76 commits made in spring 2026.

PartBuilt withSize
Web appReact 19, TypeScript, Vite26,000 lines
APINode, Express 5, MongoDB106 routes
AIGemini Live voice, Groq scoring4 personas
DataJobs, rounds, sessions, skills15 models
OpsNginx, systemd, deploy guideUbuntu 22.04

Before you offer.

What does it need to run?
Node, MongoDB, and Gemini and Groq keys. Coding rounds need a Piston runner.
What is unfinished?
An autonomous hiring agent sits on an unmerged branch.
Are these real candidates?
No, they are samples. The scores came from the product’s own AI.

Make an offer for Quasar.

Write to with your offer.

Watch the 2026 demo