Create the assignment
Paste a job description. Vettly refines it with a few clarifying questions, then generates the assignment and a weighted scoring rubric you can edit. Bring your own assignment if you prefer.
Built for hiring teams that use take-home assignments
Take-home assignments, validated by AI
Generate an assignment from your job description, invite candidates with private links, and get reports that build the code, check the logic, run a full integrity analysis, and hand you the interview questions to ask next.
Every integrity signal ships as confidence plus evidence. You make the call.
Why Vettly
Real builds, not guesses
We clone, build, and run every repository in an isolated environment before a single point is scored.
Evidence on every line
Scores cite the files and regions behind them. Integrity findings arrive as confidence plus evidence, never a verdict.
Zero candidate friction
Candidates get one private link with the brief and deadline. No accounts, no installs, nothing to learn.
How it works
From job description to decision
Invite candidates
Add emails or upload a spreadsheet. Every candidate gets a private link with the brief and deadline. No accounts, no friction.
We evaluate every submission
Each GitHub repository is built and run in an isolated environment, scored against your rubric with file-level evidence, and passed through the full integrity suite.
Decide with a real report
A composite score, stage-by-stage breakdown, a next-round recommendation, and interview questions aimed at each submission's weak spots.
The integrity suite
Six independent signals, zero accusations
Each signal is reported as a confidence level with cited evidence, so you decide with facts.
Cross-candidate similarity
Fingerprints every submission and surfaces overlapping code across candidates, with the matching regions cited.
Commit forensics
Reads the repository history for bulk single commits, implausible timing, and author mismatches.
Public repository match
Checks distinctive code against public GitHub repositories and cites what it finds.
AI-generation likelihood
Estimates how likely the code is AI-generated. Explicitly labeled unreliable and never used as a verdict.
Resume cross-check
Compares the submission against the candidate's resume claims and notes inconsistencies with evidence.
Live deployment check
If the candidate shares a deployed link, we visit it, screenshot it, and score it as a bonus. No link, no penalty.
Pricing
Simple credit packs
One credit evaluates one candidate submission. No subscriptions, no free tier, no surprises.
Starter
10 evaluation credits
₹2,499.00
₹249.90 per evaluation
Get startedGrowth
50 evaluation credits
₹9,999.00
₹199.98 per evaluation
Most popular
Get startedScale
200 evaluation credits
₹29,999.00
₹150.00 per evaluation
Get started- Isolated build and run
- Rubric scoring with cited evidence
- Six-signal integrity suite
- Interview questions per candidate
Every pack includes everything above. Launch pricing, subject to change. Payments via UPI or card, processed by Razorpay. Every purchase comes with a numbered, downloadable bill.
Stop reviewing take-homes by hand
Publish your first assignment in minutes and let every submission arrive pre-evaluated, with the evidence attached.