Every feature your hiring team needs. None of the legal risk.

No. Shynsa evaluates voice content and delivery only — communication clarity, answer relevance, confidence, fluency, and role-fit. No facial analysis, no emotion detection, no eye-tracking. Video recording, where enabled, is stored for human review only and is never fed into scoring. This matters under India's DPDP Act 2023: biometric and facial data is treated as a high-risk category requiring separate, specific, standalone consent — not a bundled checkbox — and mishandling it can attract penalties of up to ₹250 crore. Platforms that score candidates on facial micro-expressions carry that compliance burden. We don't, because we don't collect the data in the first place.

The scoring model evaluates structured dimensions separately, then combines them into an overall score. Each dimension has a written comment explaining the rating. Scoring accuracy improves with clearer question writing — specific, role-relevant questions produce more accurate dimension scores than vague open-ended prompts. Admins can always re-score any interview manually.

Yes. Questions are created per position — you write the exact text. Questions support branching conditions, thinking time (up to 60 seconds before recording starts), max answer time (10–600 seconds), and per-question follow-up prompts. There is no pre-built question library — every question is purpose-written for your role.

Branching lets you define conditions on candidate answers. For example: if the first question asks 'Are you available for night shifts?' and the candidate says no, the interview can end immediately with a polite closing message — saving the candidate's time and preventing unqualified applications from filling your pipeline. Use it for hard disqualifiers like location, availability, minimum experience, or specific certifications.

Candidates receive a link (from you, via any channel — email, WhatsApp, Naukri, LinkedIn). They open it on any smartphone browser, fill a short form (name, age, phone, email, location), accept consent, and begin the voice interview. No app download, no account creation, no interview scheduling required.

The interview resumes from where it left off. The pending question is re-read to the candidate on reconnect, and the transcript picks up without creating duplicate entries. Candidates who abandon (close the tab without completing) are marked 'abandoned' — you can see exactly how far they got.