What we collect
Account data. When you create an account we store your name, email address and authentication identifiers. Sign-in is handled by Google Firebase Authentication.
Call data. The product's job is to run phone conversations. For calls placed or received through your account we process audio streams, generate transcripts, and store call metadata such as duration, language, disposition and the custom fields you configure. This data belongs to your workspace and is used to provide the service: transcripts, analytics and billing.
Billing data. Wallet balances, ledger entries and usage records. Card and payment details are processed by Razorpay; we do not store card numbers.
Technical data. Standard server logs (IP address, request time, user agent) kept for security and debugging.
What we do not do
- We do not sell personal data.
- We do not use your call recordings or transcripts to advertise to your customers.
- We do not publish customer data, and call content is only visible inside the workspace it belongs to.
Where your data lives
Shravan.ai is an Indian company and our servers are not in India. The application, the database and stored recordings run on infrastructure in the European Union (Hetzner, Finland), with encrypted connections throughout. Call audio is carried by your telephony provider’s network while the call is in progress; transcripts, summaries and recordings are stored on our servers in the EU. If your contract requires storage in a specific region, ask us before you sign rather than assume.
Processors we rely on
Calls are carried by Plivo. Speech recognition runs on Deepgram; speech synthesis on Sarvam AI and, where configured, ElevenLabs; language models on Groq and OpenRouter. Payments run on Razorpay, authentication on Google Firebase, error monitoring on Sentry, and appointment confirmations on Google Calendar and Meta WhatsApp Business where a workspace enables them. Each processor receives only what it needs to perform its function, and nothing on that list receives your wallet, your API keys or another customer’s data. We are working through each supplier’s data-processing terms and confirming in writing how long they retain what we send and whether it is used to train their models; until that is complete for a given supplier we will not make claims on their behalf, and we will tell you the current position for any supplier you ask about. The full list, with what each one receives, is on our subprocessors page, and a signable Data Processing Addendum is available on request.
Consent for recorded calls
Every agent identifies itself as an AI at the start of the call, and where the workspace records calls it says so in the same breath. This is enforced by the platform rather than left to whoever wrote the script: a call flow whose opening omits either is refused at publish time, and the runtime speaks the missing sentence itself if one somehow gets through. Recording is off unless a workspace turns it on. You are responsible for ensuring that calls placed through your account comply with applicable law, including consent or disclosure requirements for recorded or automated calls in the jurisdictions you call into.
Retention and deletion
Recordings are kept for 180 days and transcripts for 12 months, after which they are removed automatically; the record of the call itself — its time, duration, outcome and what it cost — is kept as a billing record. A workspace can ask for shorter periods, or for longer where its own obligations require it.
You can ask us to delete your account and its data by writing to privacy@shravan-ai.tech, and we will do it within 30 days. One honest caveat: our nightly database backups are kept for a short rolling period, so deleted data survives in those backups until they rotate out. We do not restore deleted data from them, and we do not use backups for anything but recovery.
Cookies
The public site sets no cookies at all — no analytics, no advertising, no tracking pixels, nothing to consent to. Signing in to the console stores your session locally in your browser so you stay signed in; that is strictly necessary for the console to work and is cleared when you sign out. We do not use cookies to profile you or anyone you call.
Your choices, and how to use them
You can ask us for a copy of the personal data we hold about you, ask us to correct it, or ask us to delete it. Write to privacy@shravan-ai.tech; we answer within 30 days and we will tell you if something will take longer. If you are unhappy with how we handled a request, say so in a reply — it reaches a person, not a queue — and we will escalate it internally before you have to take it further.
Who is responsible for what
Two different relationships sit inside one product, and the distinction matters when someone asks for their data. For your own account — your name, your email, your billing, your support messages, our security logs — we decide what is collected and why, and you should come to us. For the people your agents call — their numbers, the call audio, transcripts and anything the agent captured — you decide what is collected and why; we process it on your instructions under our agreement with you. A caller asking to be forgotten is your request to answer, and we will carry it out on your side of the platform.
Children
Shravan.ai is built for a business to speak with the adult who contacts it. We do not provide a mechanism for verifying parental consent and we do not claim to. If your calls involve people under 18, the consent you need is yours to obtain before you configure the workspace — see our acceptable use page. We never use call data for advertising, or for profiling beyond the summary and analytics your own workspace sees.
Contact
Questions about this policy, or a request about your data: privacy@shravan-ai.tech. Anything else: hello@shravan-ai.tech. Security reports go to security@shravan-ai.tech — see the security page.
Shravan.ai is operated from Uttar Pradesh, India.