Kavach catches scam calls in Telugu and Hindi, warns the listener, alerts the family and seals the evidence on Bitcoin. Built entirely on Gnani AI: Prisma, Evon and Timbre.
The full demo is one real run of the app, narrated: a Telugu scam call, the verdict, the warning, the family alert, the 1930 complaint, the seal, a genuine call left alone, and the family asking.
This is Kavach's own screen, replaying three saved Telugu calls that it analysed with Gnani Prisma, Evon and Timbre. Press Analyse and the call plays, with English subtitles:
Live analysis needs our Gnani key and Evon on a GPU, so it runs on our machine, not here.
During the call, agents 1–7 run in a fixed order. Evon argues; the verdict is decided by code, never by the LLM.
Transcribes the call and separates the caller from the listener.
Lists the scam tactics, quoting the caller.
Lists signs the call is genuine.
Counts a claim only if the caller's own words prove it, and it isn't negated.
Warns the listener in Telugu or Hindi, and alerts the family with a voice note.
Drafts the complaint for the 1930 cybercrime helpline.
Seals the evidence in a hash chain, timestamped with OpenTimestamps.
After the call, answers the family's questions from the sealed evidence.
| 22/22 | Genuine Telugu–English calls left alone |
| 16/22 | Scams caught on the same set |
| 3/3 | Telugu calls right end to end with real Evon |
| 9–12 s | Time to a Telugu verdict |
| 120 | Automated tests passing |
| Prisma v2.5 | Speech-to-text in Telugu, Hindi, English and code-mixed speech; separates caller and listener |
| Evon v3.3 | Open-weight LLM, self-hosted |
| Timbre | Speaks the warning and the family's voice note, in Telugu and Hindi |