deepidv Logo
Verify Humans
with your AI Agent

Your agent can now verify any human they transact or interact with. One command to start.

Launch with Claude
Launch with GPTLaunch with GeminiLaunch with GrokLaunch with Meta
or copy the prompt below
Hi Claude! Let's launch deepidv for my AI agent. Research the docs at https://mcp.deepidv.com and get it set up.
Single Command

Your agent sends a single request

ID document, selfie, or verification link — one API call triggers the full pipeline.

Verification Skills

Agent + Human Verification

Automate your process, protect your autonomous agent and embrace TL-Encryption verifications.

Document Authentication

Agent sends a photo of any ID — gets a full forensic breakdown across 6,500+ document types in 195 countries. Classification, extraction, and tamper detection in one call.

Biometric Face Match

Trigger face comparison between a selfie and document photo. 512-dimensional embeddings trained on 12M+ pairs deliver sub-100ms matching with industry-leading accuracy.

Deepfake Shield

Every submission is screened for AI-generated identity attacks. Ensemble detectors target frequency artifacts, blending boundaries, and physiological implausibility in real time.

Passive Liveness

Single-frame liveness analysis runs in parallel with biometrics — no blinking, no head turns. Invisible to the user, impossible to spoof with static images.

Risk Intelligence

Six specialized AI agents — Device, Velocity, Geolocation, Behavioral, Network, and Document Forensics — score every session. Fully explainable, never a black box.

TripleLock Encryption

End-to-end encrypted verification results. Three independent key holders must cooperate to decrypt — AES-256 at rest, TLS 1.3 in transit. Zero plaintext exposure.

Universal Compatibility

Any Verification. Any Provider.

Our MCP server speaks one protocol that every modern agent already does — Claude, GPT, Gemini, Grok, Llama, Cursor, Copilot, Perplexity, your in-house model. Hook the deepidv verification engine into whichever agent your team runs today, and switch tomorrow without rebuilding a single integration.

Example Prompts

Test your connection with these

Once your agent is connected, drop any of these prompts in to see the deepidv MCP server respond.

Verification Sessions

  • List the most recent 5 verification sessions
  • Show me the details of verification session [session_id]
  • Create a verification session for John Doe, email john@example.com, phone +1234567890
  • Get the artifacts for verification session [session_id]
  • Mark verification session [session_id] as verified
  • Reject verification session [session_id]

Bank Statements

  • List all bank statement requests
  • Show me the bank statement for [statement_id]
  • Find bank statements linked to external ID [your_id]
  • Create a bank statement request for Jane Smith, email jane@example.com, phone +1234567890

Workflows

  • List all available verification workflows
  • Show me the details of workflow [workflow_id]

General / Exploratory

  • What tools do you have access to from deepidv?
  • Walk me through the status of the last 3 verification sessions
  • Are there any verification sessions created today?

MCP & Agent Verification FAQ

Basically anything. If your agent can run a tool in a command shell, it works with deepidv. Claude, GPT, Gemini, Grok, Meta AI, open-source models — any agent that supports MCP or can execute shell commands can trigger verifications.

Add the deepidv MCP server to your agent's configuration — it takes one command. Once connected, your agent can call verification tools directly. Full setup instructions are at mcp.deepidv.com.

Yes. Every verification is end-to-end encrypted with our patent-pending TripleLock system — three independent key holders must cooperate to decrypt biometric data. All data is AES-256 at rest, TLS 1.3 in transit, with zero plaintext exposure at any point in the pipeline.

No. MCP access is included in your deepidv subscription at no additional cost. Same pricing, same API limits — you're just giving your agent the ability to call the same verification engine you already have access to.

MCP (Model Context Protocol) is an open standard that lets AI agents connect to external tools and services. Think of it as a universal plug that lets any AI model call deepidv's verification engine directly — no custom integration code required.

Every verification evaluates 50+ forensic features per document across 6,500+ document types in 195 countries. That includes document authenticity, biometric face matching, passive liveness detection, deepfake screening, and real-time behavioral risk scoring from six specialized AI agents — all in a single call under 150ms.

Still have questions?

Our team is ready to help you get started.

Contact Sales

Give your agent the power to verify

One command connects your AI agent to the deepidv verification engine. Included in every subscription.

Read the Docs