On this page
Crivello is built and tested on our machines. It is not live yet, and it is not a registered company yet. This is a draft for a lawyer to review; we do not claim compliance with any law yet.
1. Who we are#
In short. Crivello is one founder in Ludhiana. For most data, we decide how it is used.
Crivello (“we”) runs crivello.ai and the Crivello app. Crivello is not a registered company yet. It is run by its founder, Karanjot Singh, in Ludhiana, Punjab, India. Write to content@crivello.ai.
For data you give us directly, we decide how it is used. India’s DPDP Act calls this a data fiduciary; the GDPR calls it a controller. For candidate data we handle for a hiring company, that company also decides how it is used.
2. Who this covers#
In short. Visitors, founders who hire, and candidates who interview.
- Visitors to crivello.ai.
- Founders and hiring teams who use Crivello to hire.
- Candidates who apply to, or interview for, a role that uses Crivello.
3. What we collect#
In short. What you type, what you say in an interview, and a few numbers about the interview room. Never your ID images.
| Kind | What |
|---|---|
| Account | Name, email, company, and your password, stored hashed |
| Applications | Resume and the answers you give when you apply |
| Interviews | Video and audio recordings, transcripts, scores, the quotes behind each score, reviewer notes |
| Interview room activity | Leaving the window, pasting, typing bursts and code that matches an AI answer. Sent as events and numbers, never as key presses |
| ID check, only if the role asks | Pass or fail, the document’s country and type. Never your ID images, name, date of birth or document number |
| WhatsApp, only if you opt in | Your phone number, for updates about your application |
| Payments | Handled by Dodo Payments. We do not hold card details |
| Website | Pages visited and device data, through analytics (part 12) |
| Messages | What you send us by email or the contact form |
Camera checks are building next. They are switched off, so we collect nothing from them.
4. Why we use it#
In short. To run hiring, to check our AI, to stop cheating, to bill, and to answer you. Nothing else.
| Why | Basis |
|---|---|
| Run applications and interviews | Your consent, asked before each AI round (AI interview, recording, data processing); our contract with the hiring founder |
| Score interviews and check the AI | Your consent; under the GDPR, also our legitimate interest in checking quality |
| Note activity in interview rooms | Your consent, asked before each room, with a notice of what is noted |
| Billing | Our contract with the founder, and legal duties |
| Answer messages | Your consent |
| Keep the service safe | Legitimate interest (GDPR); a legitimate use the DPDP Act allows |
Under India’s DPDP Act we rely on your consent, or on a legitimate use the Act allows. Under the GDPR we name one basis per use, as in the table.
5. AI and automated decisions#
In short. AI screens resumes and scores interviews. A screen-out comes with a reason, and you can reach a person. The company decides who it hires.
AI reads each resume against the role. A resume that hits a knockout, or clearly misses a must-have skill, is screened out automatically. You get a written reason, and you can reply to reach a person. If a resume is unclear, it counts as a yes, so thin resumes still get an interview.
AI scores interviews. Every score must quote your own words, with the time in the interview. A second AI model, from a different company, re-scores each interview without your name. When they disagree on a must-have skill, a person decides. A person on our team reviews every finalist packet. The hiring company makes the hiring decision; its final round is with a person, and we do not score it.
You can ask a person to re-check any result. Under the GDPR this is your right not to be subject to a decision based only on automated processing.
6. Who sees it#
In short. The company you applied to, our review team, and the providers that run Crivello. We do not sell your data.
- The hiring company sees its own applicants and finalists. It may share your interview with a co-founder involved in the hiring decision.
- Our review team sees interviews to check them.
- The providers in part 7, only to run Crivello.
- Authorities, where the law requires it.
We do not sell your data.
7. Providers we use#
In short. AI models, video calls, email, payments and hosting come from other companies. The list is not final.
These are the providers our code is built for today.
| What it does | Provider |
|---|---|
| AI models for screening, interviews and scoring | Anthropic, Google and OpenAI models through Vercel AI Gateway; Azure AI Foundry |
| Speech to text and voice | Deepgram, Cartesia or ElevenLabs, depending on setup |
| Video calls and recordings | Daily |
| ID check, only if the role asks | Didit |
| Payments | Dodo Payments |
| Resend | |
| Background jobs | Inngest |
| File storage | S3-compatible storage |
| Hosting and website analytics | Vercel, Google Analytics, PostHog and Microsoft Clarity, from page load |
8. Data that leaves India#
In short. Some providers work outside India, for example in the United States or the EU.
Some providers handle data outside India and outside your country. We choose providers that offer contract terms to protect it.
9. How long we keep it#
In short. Recordings 180 days after the interview. Transcripts and scores one year. Then they go.
| Data | How long |
|---|---|
| Interview recording | 180 days after the interview, or sooner when you ask |
| Transcript, feedback, resume and scores | 365 days |
| Consent records | 7 years |
| Backups | Age out within 35 days |
| Messages to us | 12 months |
The hiring founder may set shorter times. After you ask us to delete everything, we keep only statistics that cannot be traced back to you, and a log that you asked, with the date and nothing about you. A restored backup deletes your data again before use.
10. Your rights#
In short. See, fix, download or delete your data, and take back a consent. We answer within 30 days.
- See what we hold about you, and download it.
- Correct it, or delete it.
- Withdraw a consent. This does not undo what happened before you withdrew it.
- Ask a person to review an AI result.
- Under the DPDP Act: name someone to act for you if you cannot, and use our grievance officer (part 14).
- Under the GDPR: also limit how we use your data, object to it, and take it to another service.
Candidates: use the data page linked in your invite email. Anyone: write to content@crivello.ai. We confirm we got your request within 3 business days and answer within 30 days.
11. Children#
In short. Crivello is for adults, 18 and older.
Crivello is for people 18 and older. We do not knowingly collect data from children. If you think a child has given us data, write to us and we delete it.
12. Cookies and analytics#
In short. Analytics load with the page. Press Reject to turn them off.
| Tool | Cookies |
|---|---|
| Vercel Analytics | None |
| PostHog | Yes, from page load; Reject turns it off |
| Google Analytics | Yes, from page load; Reject turns it off |
| Microsoft Clarity | Yes, from page load; Reject turns it off |
Change your choice any time with Cookie settings in the footer.
13. Security#
In short. Passwords are hashed, and only people who need data can see it.
Passwords are stored hashed. Only people who need data to do their job can see it. The Security page shows how candidate data is kept safe, and what is built and what is not.
14. Grievance officer and complaints#
In short. Write to Karanjot Singh first. You can also go to the regulator.
Grievance officer: Karanjot Singh, content@crivello.ai. If our answer does not settle it, you can complain to the Data Protection Board of India, or, in the EU or UK, to the data protection authority where you live.
15. Changes#
In short. If a change matters, we email you first.
If a change matters to you, we email you before it applies. The date at the top shows the last change.
16. Contact#
In short. content@crivello.ai.
Crivello, Ludhiana, Punjab, India. content@crivello.ai.
