FinHub Verify · AI
Liveness, face match and deepfake detection from a session the customer records themselves — for the onboarding journeys that do not require a live official on the call.
The Challenge
Agent-led verification does not scale past the agents
Every session needs a trained official and a slot in their day, so throughput is capped by headcount and by business hours.
Customers drop off in the wait. An application that needs an appointment tomorrow is competing with one that finishes tonight.
Not every journey legally needs a live official — but if agent-led video is the only route you have, all of them get treated as though they do.
Capabilities
What runs in an automated session
The checks a good agent would make, plus one they cannot make by eye.
Liveness, without a person watching
The customer records their own session. Liveness checks confirm a live human is present rather than a photograph, a replayed video or a screen held up to the camera.
- Presence confirmed from the capture itself
- Face matched against the identity document provided
- Runs whenever the customer is free, not when an agent is
Deepfake detection built in
Synthetic video is now cheap enough to be a real onboarding threat. Detection runs on every session rather than being reserved for cases somebody already found suspicious.
- Synthetic and manipulated video flagged during verification
- Applied to every session, not a sampled subset
- Flags surface for review rather than silently failing the customer
Onboarding that does not wait for a queue
The drop-off in agent-led video verification is largely queueing and scheduling. Removing the appointment removes the wait, which is where applications are lost.
- No agent scheduling and no queue
- Available outside business hours
- Branch verification cost and overhead removed for these journeys
Drops into the flow you already have
Verification is one step inside an onboarding journey you have already built. It connects to the platforms running that journey rather than asking you to rebuild it.
- Integrates with existing onboarding platforms
- Sits alongside the verification APIs on the same account
- Consent captured and written to a tamper-evident audit trail
Choosing
Intelli VKYC or agent-led V-CIP?
They are not competing options. The regulation decides which one a given journey needs.
| Criteria | Intelli VKYC | Video KYC (V-CIP) |
|---|---|---|
| Who runs the session | The customer, alone | A trained official of the regulated entity |
| Satisfies RBI V-CIP | No | Yes, when run to the Master Direction |
| Availability | Any time | Business hours and agent capacity |
| Scales with | Volume | Headcount |
| Typical use | Journeys where V-CIP is not mandated, or a pre-check | Onboarding in V-CIP scope |
Which of your journeys fall inside V-CIP scope is a question for your compliance function against the current RBI Master Direction on KYC, not for a vendor page.
How a session runs
Start to sealed record
Five steps, none of which require an appointment.
Customer starts the session
From your app or a link, on their own device and in their own time. No appointment, no queue.
FAQ
Intelli VKYC — common questions
What it verifies, what it catches, and where V-CIP still applies.
AI-led video KYC verifies a customer from a self-recorded video session rather than a live call with an agent. Liveness detection confirms a real person is present, face match compares them against their identity document, and fraud checks look for synthetic or manipulated video. The customer completes it whenever they choose, with no appointment.
Explore More
Related verification
Run it on your own onboarding flow
A pilot on a journey you pick, so you can measure completion and drop-off against your current process.