New:Ravan Revenue Recovery — voice AI priced on recovery, not on calls →
Trust Center

Built for secure, auditable voice operations.

Agni is designed for teams that need control over customer conversations, call data, access, and operational review. Every call is recorded, transcribed, and reviewable — and India's calling rules are built into the workflow, not bolted on afterwards.

Access controlsAudit-ready call recordsHuman handoffData handling controls
Operational controls

Controls your security team can verify.

Nothing about an AI phone call should be a black box. Agni gives administrators the access controls, records, and review workflows to see exactly what every agent said and did.

Workspace access & team permissions

Control who can build agents, launch campaigns, and review call data. Keep production calling separate from experimentation.

Complete call records

Every call produces a recording, a full transcript, and session details — duration, disposition, disconnect reason, and cost. Nothing is summarized away.

Configurable data flows

Decide where call data goes. REST APIs and webhooks deliver transcripts, recordings, and outcomes to the systems you choose — see supported integrations.

Human handoff for sensitive calls

Define when agents must transfer to a person — disputes, vulnerable customers, regulated disclosures — with conversation context attached.

Post-call review workflows

Review transcripts, sentiment, and outcomes call by call. Flag, audit, and correct agent behavior before the next campaign run.

Safe testing before deployment

Test agents with web calls before they ever dial a real customer. Validate prompt and script changes against real scenarios first.
India compliance

India calling compliance, built into the workflow.

Most platforms treat India's calling rules as your problem to solve after deployment. Agni treats them as workflow requirements — scrubbing, templates, windows, disclosure, and records are part of how campaigns run.

01

DND/NCPR scrubbing before dialing

Contact lists are screened against Do-Not-Disturb preferences before a campaign dials — not after a complaint.

02

DLT template workflows

Operate within 140-series (promotional) and 160-series (service) calling contexts, with campaigns aligned to your registered DLT templates.

03

Consent logging

Capture and store the consent basis for each contact, so every dial can be traced back to a recorded opt-in.

04

Calling-window enforcement

Campaigns respect configured calling hours — including RBI-aligned windows for collections workflows — automatically.

05

AI-call disclosure support

Configure agents to disclose that the call is AI-assisted at the start of the conversation, in the customer's own language.

06

Per-campaign audit trails

Every campaign keeps complete records — who was called, when, under which template, and what was said — ready for internal or regulator review.

What the law requires — a short primer

If your team runs outbound calling in India, two regimes shape the work. Here is what they ask for, so you can map Agni's workflow controls against your own obligations.

Commercial calling

TRAI TCCCPR & DLT

  • Businesses (principal entities) and their telemarketers must register on DLT platforms, with PE–telemarketer linkage.
  • Promotional calls run on 140-series numbers; service and transactional calls on the 160-series.
  • Call scripts follow pre-registered DLT templates, and contact lists must be scrubbed against DND/NCPR preferences.
  • Commercial calls are restricted to a 10 AM – 7 PM window, and AI-generated calls must be disclosed.
  • Penalties range from ₹1,000 to ₹1.5 lakh per violation; repeated complaints can trigger disconnection.
Source: TRAI TCCCPR amendments, gazette Feb 2025
Personal data

DPDP Act 2023

  • Rules notified in November 2025, with a full-compliance deadline of May 13, 2027.
  • Consent notices must be itemized — and available in English plus India's scheduled regional languages.
  • Bundled consent is not permitted, and withdrawing consent must be as easy as giving it.
  • Organisations need data processing agreements governing the sub-processors that handle personal data.
  • Penalties for breaches run up to ₹250 crore.
Source: DPDP Act 2023 & DPDP Rules, notified Nov 2025

Running lending or collections workflows? RBI's Digital Lending Directions hold regulated entities accountable for the conduct of their agents — including AI agents — and restrict recovery contact hours. Agni's calling-window enforcement and complete call records are designed to support those controls. For the full clause-by-clause mapping, download the RBI Compliance Pack on Ravan Revenue Recovery, or model the economics in the ROI calculator.

This summary is for orientation only and is not legal advice. Your obligations depend on your registration status, industry, and use case — Agni provides the workflow controls; your compliance program defines how you use them.

Data protection

Designed for DPDP-era consent management.

The DPDP Act changes how Indian businesses collect, use, and retain customer data. Agni's calling workflows are designed so consent, withdrawal, and review fit how the law expects data to be handled.

Readiness posture

  • Itemized consent design — campaigns can be scoped to the specific purpose a customer agreed to
  • Withdrawal flows — opt-outs on a call are recorded as structured outcomes, and contacts can be suppressed from future campaigns
  • Data principal rights support — complete recordings, transcripts, and session details make access and correction requests answerable
  • Retention and deletion terms documented per workspace as part of enterprise onboarding
  • DPAs with sub-processor terms available on request for enterprise agreements

Consent in the customer's own language

DPDP requires consent notices in English plus regional languages. Most platforms treat that as a translation chore. Agni speaks 30+ Indian languages and handles 100+ accents & dialects — so consent, AI disclosure, and opt-out conversations can happen natively in the language your customer actually speaks.

30+ Indian languages100+ accents & dialectsHinglish native
Certifications

Where we are on certifications, honestly.

Our architecture is SOC 2, GDPR, and HIPAA ready — controls, data handling, and agreements built to those standards — with formal certifications in progress rather than implied. Enterprise customers can request our current security documentation and DPA today, and we complete security questionnaires as part of evaluation.

SOC 2SOC 2 ready — Type I certification in progress
GDPRGDPR ready — DPA and transfer safeguards available
HIPAAHIPAA-ready deployment options for healthcare workflows
ISO 27001On certification roadmap
Security documentation & DPAAvailable on request
Responsible AI

Responsible AI calling, as policy.

Customers should never feel tricked by an AI call. These are the defaults we design for — and the standards we recommend every team configures.

AI disclosure

Agents can identify the call as AI-assisted at the start of the conversation, in the customer's language.

Human escalation

A path to a person is always configurable. For sensitive workflows, we recommend making it mandatory.

Recording transparency

Calls are recorded for quality and audit. Agents can state this wherever your policy or applicable law requires it.

Grievance contact

Concerns about a call from an Agni-powered workflow can be raised at info@ravan.ai. We route them to the operating business and to our own review.
Procurement FAQ

Questions security teams ask us.

The short answers, before the questionnaire. For anything not covered here, our team responds directly — no portal, no ticket queue.

DND/NCPR scrubbingConsent loggingCall recordings & audit trails

Evaluating the rest of the platform? See pricing, integrations, the ROI calculator, and the general FAQ.

Where is call data stored? Do you support India data residency?
We offer deployment options including India data residency for enterprise engagements. Residency posture depends on the deployment model and telephony configuration, so we scope it per engagement — contact us with your requirements and we'll document the architecture for your review.
How long are call recordings and transcripts retained?
Retention is agreed per workspace as part of onboarding. Recordings, transcripts, and session details stay available for your audit and review workflows, deletion requests are supported, and enterprise agreements document retention and deletion terms in the DPA.
Do you use sub-processors?
Yes — like most cloud platforms, Agni relies on a small set of infrastructure and telephony sub-processors. The current list, and the data processing terms that govern them, are available on request as part of enterprise evaluation.
Will you sign a DPA?
Yes. We provide a data processing agreement covering processing roles, sub-processors, and data principal rights support — and we can review your standard paper for enterprise agreements. Email info@ravan.ai to start.
Can we send you our security questionnaire?
Please do. Send it to info@ravan.ai and we'll complete it as part of your evaluation, alongside our current security documentation. If you'd rather walk through it live, book a call.

Bring your security questionnaire.

Send your questionnaire, DPA, or audit requirements to info@ravan.ai — or book a demo and walk through Agni's compliance workflows live with our team.

or call our AI agent directly+91 11 4056 6601Available 24/7 · Hindi, Hinglish, EnglishChat on WhatsApp