What Margin keeps, and why
Margin is an interview practice room run by a small founding team in India. This notice explains what we collect when you practise, who processes it for us, how long we keep it, and how to have it deleted.
- No account is needed. A random ID in a cookie links your reports and resume preparations to this browser.
- Your interviews and resume text are processed by Google's Gemini models through Google Cloud.
- We don't sell your data, show ads, or use it to train AI models.
- You can delete any report or resume preparation in the app, or email us to delete everything.
What we collect
| Data | Why | How long |
|---|---|---|
A random browser ID in the margin_owner cookie | To show you only your own reports and preparations, and to apply daily limits | 6 months from your first visit |
| Your interview: what you type or say (as a transcript), and your code, queries, notes, canvas and test results | To run the interview and write your report | The live session ends after 60 minutes; the report keeps the parts it quotes |
| Voice audio, when you choose voice mode | Streamed to Gemini so the interviewer can hear you. Margin does not store the audio | Not stored by Margin |
| Interview reports | So you can review them and practise again | Until you delete them |
| Resume text, the role and job description you add, your answers to clarifying questions, and public pages you select | To prepare a resume interview. The uploaded file itself is discarded once its text is extracted | Deleted automatically after 90 days, or sooner if you delete the preparation |
| Feedback you send about a report, and an email address if you choose to give one | To improve Margin and reply to you | 12 months |
| Anonymous usage events (pages visited, interviews started and finished), tagged with a one-way hash of the browser ID | To understand how people use Margin. No names, emails or interview content | 30 days in our logs |
Who processes it
Google Cloud hosts Margin and stores reports and preparations in Mumbai, India. Google's Gemini models on Vertex AI run the interviewer, write reports and read resumes. These models may process your data outside India; today the interviewer's voice model runs in the United States and the evaluation models run on Google's global endpoints. Google acts as our service provider under its Cloud terms, which don't allow it to use this data to train its models.
Code you run in Python or SQL exercises runs in your own browser. The Python runtime loads from the jsDelivr content network.
Your choices and rights
You can delete a report from My reports and a resume preparation from its page at any time. Under India's Digital Personal Data Protection Act, 2023, you can also ask us to access, correct or erase your data, or withdraw consent. Write to us and include a report link from your browser so we can find your data, since we don't know your name.
Margin is meant for people aged 18 and over.
Contact
Email aumvyas0202@gmail.com with questions, deletion requests or complaints. We reply within 7 days.
If we change how we handle data, we'll update this page and the date above.