eConsent Voice

Voice consent, captured and defended.

Real-time consent classification, offshore agent tracking, and evidence for dispute defense for AI voice agents, BPOs, and in-house dialers.

Capture
Verbal consent
Verify
Before dial
Export
Certificate

How it works

Check, capture, certify - in one call

One classifier, one signing key, one record. As the call moves from precall check to live consent to signed evidence, every decision and disclosure lands in the same certificate — scored, timestamped, and hashed.

1. Precall
Check consent before the dialer connects
Call state, disclosure state, and revocation state are checked together.
2. Live call
Capture permission while the call is still active
Transcript markers and consent labels land in the same record as the call.
3. Evidence
Export a certificate with audio and disclosure hashes
The output is built for audits, partner packets, and dispute defense.
Call view
Voice consent detail view
Transcript label
"Yes, I agree to receive calls about solar panel installation from SunPower."
AFFIRMATIVE
Latency
<200ms
Rules and model inference stay inside the call path.

Offshore Compliance

Built for proposed FCC offshore-agent rules

When calls involve offshore agents, proposed FCC rules would require explicit disclosure and consent transfer. eConsent Voice tracks every agent in the call chain with location, role, and handoff timestamps so you have the record if and when those rules take effect.

Agent Chain Detail
Offshore agent chain tracking

Compliance checks in every certificate

Jurisdiction risk flag
Flags calls where agent location may trigger additional disclosure obligations so your compliance team can review before proceeding
Offshore Location Disclosure
Detects and records offshore caller disclosure in the transcript; flags certificates missing the disclosure
Right-to-transfer notice
Detects whether the consumer was informed of their right to a US-based representative, to align with the FCC's March 2026 NPRM
Disclosure Timeline
Every required disclosure timestamped and baked into the certificate

Evidence Comparison

You already have consent data.
Is it defensible?

A side-by-side look at what your dialer exports today versus the evidence bundle eConsent Voice produces for every call.

Without eConsent
dialer_export.csv
- phone: +1 (555) 000-1234
- timestamp: 2026-04-10 14:32:00
- consent_text: "by clicking submit"
- source: spreadsheet_row_428
- ip_address: (not captured)
- audio: (none)
- transcript: (none)
- disclosures: (none)
- signature: (none)
Plaintiff's counsel will argue
No audio, no signature, no disclosure log, no proof the consumer ever heard the terms.
With eConsent Voice
consent_bundle.json
+ consent_id: cst_mnwd5l1p...
+ phone_hash: 84d12ceb...
+ captured_at: 2026-04-10T14:32:00Z
+ uploaded_at: 2026-04-10T14:32:47Z (server-signed)
+ transcript: "Yes, I agree..." (227 words)
+ audio_sha256: 7abfe60b...
+ audio_storage: s3://econsent-voice-audio/ (Object Lock COMPLIANCE, 7yr)
+ classifier_verdict: accept_consent (0.988)
+ disclosures_detected: 7/7 FCC required
+ signature: HMAC-SHA256 (2430265984b6c2ad...)
What a defensible record looks like
Cryptographic chain-of-custody from recording to certificate. Every field verifiable independently.

Platform-agnostic

One precall check, any voice platform

One REST API sits in front of the platform you already dial from. Any stack that can make an HTTP request - AI voice agents, cloud dialers, or contact center suites - calls the check before the conversation starts.

Vapi Retell Bland AI Synthflow Ringba Five9 NICE Genesys

Examples of platforms teams already run. eConsent sits in front of any of them with a single HTTP call — no formal integration or partnership required.

Developer API

Four API calls. Full compliance chain.

Every outbound call passes through a four-stage pipeline that captures, classifies, and certifies consent in real time.

Verify consent before every dial

Before your agent connects, send a precall check. The API validates consent status, checks revocation history, verifies campaign permissions, and returns a go/no-go decision with the exact FCC disclosure script - in under 200ms.

Phone number, campaign ID, agent ID required
Returns allow, review, or deny
Consent score from 0-100 included
// POST /v1/precall
{
"phone": "+15551234567",
"calling_entity": "SunPower Solar",
"purpose": "marketing",
"call_type": "ai"
}
// Response
{
"allowed": true,
"decision_id": "dec_8f3k2m",
"consent": {
"consent_id": "cst_a1b2c3",
"type": "prior_express_written",
"ai_specific_consent": true
},
"required_disclosures": ["caller_identity", "recording_notice"]
}

Pricing

Voice pricing

Pay only for the voice events you use. Start with $25 credit and a monthly free tier. Enterprise custom pricing available at scale.

Free

$0

$25 starting credit

50 precall checks/mo

10 verbal captures/mo

  • All compliance features
  • Single seat
  • No credit card

Pay as you go

Most popular
$0/mo + usage

Precall check: $0.05

Verbal capture: $0.30

Volume discounts at scale

  • Whisper transcription + speaker diarization
  • S3 Object Lock 7-year archival
  • Webhook delivery + audit log
  • Multi-seat dashboard

Enterprise

Custom
Custom

Reserved capacity

Custom voice rates

Annual invoicing

  • Voiceprint fraud detection
  • Priority transcription queue
  • DPA + custom MSA
  • SSO / SAML
  • White-label certificate PDFs

Volume discounts apply automatically. See the full rate card at /pricing/rates.

Frequently asked questions

Add a precall HTTP request before your agent connects. Send the phone number, calling entity, purpose, and call type. The API returns an allow/deny decision with any required disclosure scripts. During the call, your transcribed consent is captured and classified. After the call, the evidence bundle is finalized. We provide SDKs for Node and Python, or use any HTTP client.
When the classifier's verdict is review_consent, the consent is not accepted outright - no valid certificate is issued. The request returns the transcript, classifier scores, and reasoning so your workflow can trigger a supervisor review, re-ask the consumer, or route the call differently.
eConsent Voice is designed around key FCC actions affecting outbound calling: the February 2024 declaratory ruling classifying AI-generated voices as "artificial" under the TCPA and requiring consent, the consent-revocation rules in effect since April 2025, and the March 2026 NPRM on offshore agent disclosure. Certificates capture the evidence typically needed to demonstrate consent and required disclosures. eConsent does not provide legal advice - defense strategy and admissibility remain decisions for your counsel.
By default, the classifier runs server-side on our infrastructure with quantized ONNX models for deterministic latency. On Custom plans, we offer on-premise deployment options where the model runs within your infrastructure for data sovereignty requirements.
7 years using AWS S3 Object Lock in COMPLIANCE mode. Every certificate is SHA-256 hashed at creation. The hash, along with all evidence data, is immutable and cannot be altered by anyone including eConsent. This exceeds the 4-year TCPA statute of limitations.
Yes. The API is platform-agnostic. Add precall checks in your agent's server-side function, send disclosure events during the conversation, capture consent from the transcript, and finalize after the call ends. Any system that can make HTTP requests works.

Get started today

Start verifying consent before your next dial

Get your API key in under 5 minutes. 14-day free trial with 100 precall checks. No contracts.

See eConsent in action. Schedule a live demo
Schedule a demo