We benchmarked every major assessment platform.
HackerRank. Codility. Mercer Mettl. TestGorilla. iMocha. Here's what the data shows.
Data sourced from published documentation. Last verified March 2026.
THREE WHITE SPACES
Capabilities that don't exist in the market.
Cryptographic Evidence Chain
PATENT PENDINGWHAT EXISTS IN THE MARKET
No assessment platform offers cryptographic signing of records. Scores are stored in databases that can be modified without detection.
WHAT VERIL.AI DOES
Ed25519 digital signatures + RFC8785 canonicalization. 4-layer hash chain. Anyone can verify independently without contacting Veril.ai.
WHY IT MATTERS FOR RECRUITERS
You get mathematical proof the score hasn't been altered. No trust required.
Real-Time AI Question Generation
PATENT PENDINGWHAT EXISTS IN THE MARKET
Every competitor draws from a pre-built question bank. HackerRank: 7,500+. Codility: 2,800+. All leakable, all circulating on forums.
WHAT VERIL.AI DOES
Zero question bank. Every question generated uniquely by AI — specific to the candidate's skills, level, and this exact session.
WHY IT MATTERS FOR RECRUITERS
The score reflects actual ability, not access to leaked material.
Candidate-Paid Micro-Assessment
WHAT EXISTS IN THE MARKET
Assessment platforms charge recruiters $5,000–$50,000/year. Candidates are passive participants — tested without ownership.
WHAT VERIL.AI DOES
Candidates pay ₹30 per assessment. Own their proof forever. Share it anywhere. Recruiters pay only to browse verified talent.
WHY IT MATTERS FOR RECRUITERS
Pre-verified talent pool at a fraction of the cost. No wasted assessment spend.
FULL CAPABILITY MATRIX
8 capabilities. 6 platforms. The data speaks.
| Capability | Veril.ai | HackerRank | Codility | Mettl | TestGorilla | iMocha |
|---|---|---|---|---|---|---|
| Keystroke Dynamics | ◉ | ◐ | ✗ | ✗ | ✗ | ✗ |
| Gaze Tracking | ◉ | ◐ | ✗ | ◐ | ✗ | ✗ |
| Code Progression Analysis | ◉ | ◐ | ◐ | ✗ | ✗ | ✗ |
| Integrity Score (Fused Multi-Signal) | ◉ | ✗ | ✗ | ✗ | ✗ | ✗ |
| Question Generation | ◉ | ✗ | ✗ | ✗ | ✗ | ✗ |
| Cryptographic Record Verification | ◉ | ✗ | ✗ | ✗ | ✗ | ✗ |
| Portable Candidate Credential | ◉ | ✗ | ✗ | ✗ | ✗ | ✗ |
| Integrated Job Marketplace with Proof | ◉ | ✗ | ✗ | ✗ | ✗ | ✗ |
Data sourced from: HackerRank, Codility, Mettl, TestGorilla, iMocha documentation. Last verified March 2026.
PROCTORING DEEP-DIVE
How proctoring actually works.
HOW EVERYONE ELSE DOES IT
Event log of suspicious activities. Human review queue. Noisy signals with high false positive rates. Binary pass/fail decisions.
HOW VERIL.AI DOES IT
5 behavioral signals captured simultaneously and fused into one weighted Cognitive Authenticity Score (CAS). Patent-pending fusion algorithm with confidence levels. Override rule: any single signal below 20% triggers automatic flag.
All 5 signals within normal parameters. High confidence in assessment authenticity.
Minor deviations detected but within acceptable bounds. Assessment is valid.
Significant deviation in one or more signals. Manual review recommended.
Critical integrity failure. Assessment result should not be relied upon.
FOR RECRUITERS — BOTTOM LINE
What this means for your hiring.
Scores you can mathematically verify — not just trust
Candidates who proved their skills under patent-pending proctoring — not just claimed them
Evidence that travels with the candidate — not locked in a vendor dashboard