Sector guardrail templates ship with the platform, so your agent starts from a compliant script instead of a blank prompt.
Lending & collections
Guardrail scripts for loan servicing, EMI reminders, and recovery calls — disclosure language, no harassment patterns, within-hours contact only.
Insurance
Templates for policy servicing, renewal reminders, and claim-status calls — benefit illustrations and exclusions stated as scripted, never improvised.
Mutual funds & securities
Guardrails for investor-facing calls — risk disclosures read on every relevant call, no return guarantees, scripted handling of advice-adjacent questions.
These are enforcement templates that keep agent behavior inside your compliance script — they are not an RBI, IRDAI, or SEBI certification of your operations. Your compliance team owns the rules; the platform enforces them on every call and logs the evidence. Details on Compliance.
TRAI's rules aren't a settings page — they're a pre-dial gate every outbound call passes through before it's placed.
TRAI's 9 AM–9 PM IST commercial calling window is encoded in the pre-dial policy engine — calls outside the window aren't placed, across campaigns and automations alike.
A single suppression list across voice, WhatsApp, and email. Spoken opt-outs are detected on call transcripts in English and Hindi — and it's fail-closed: if suppression can't be verified, the call isn't placed.
Explicit DLT-verified consent, inferred, and service-context tiers — each dial decision is evaluated against the consent evidence on file before it goes out.
Every call decision is recorded with its rule evaluation, reasons, and consent evidence in an append-only ledger — built to survive an enterprise or regulatory audit.
Shadow mode evaluates and logs every dial decision without blocking — collect audit evidence before you enforce. See the compliance engine and AI voice agents in India.
EMI reminders and recovery calls that stay polite, scripted, and inside permitted hours — every commitment the customer makes is captured and logged against their account.
Within-hours dial
Policy engine clears the call for the 9 AM–9 PM window and verified consent.
Scripted conversation
Amount, due date, and options stated per the Fair Practices template — in the customer's language.
Promise-to-pay logged
The commitment is written back to your CRM with the transcript attached.
Opt-out honored instantly
A spoken "stop calling me" suppresses the customer across every channel.
The agent never hears or speaks a card number. Customers complete payment on the provider's own PCI-compliant checkout, so card data never enters call audio or transcripts.
When a customer agrees to pay, the agent sends a payment link over WhatsApp or SMS — during or after the call — using Stripe, PayPal, or Razorpay connectors. The customer pays on the provider's checkout; your transcripts never carry card data.
Payment status can trigger the next step automatically — a receipt confirmation, an updated follow-up sequence, or a CRM write-back — through Automations.
Onboarding funnels leak at paperwork. Voice + WhatsApp follow-ups recover applicants before they go cold — without a manual calling team chasing every file.