Language, telephony, payments, and residency, engineered for India from the start.
Sarvam across STT, LLM, and TTS, Hinglish code-switching support and Indian-accented voices that sound local, not imported.
Lakh/crore number normalization, BFSI keyterm boosting, and multilingual tone detection built for how India actually speaks.
Vobiz (Indian CPaaS) alongside global providers, with India-originating call handling for compliant commercial routes.
Razorpay billing built in, INR invoicing and GST handled at checkout, not bolted on.
Business data lives in the client's own database on Indian infrastructure, it never has to leave India.
Full WhatsApp Business automation, the channel Indian consumers actually live on.
Every outbound call is evaluated against the encoded ruleset before it's placed, across AI Agent campaigns and pre-call automations alike.
The TCCCPR ruleset ships as an executable policy, not a PDF. The 09:00–21:00 IST commercial calling window, 140-series (promotional) vs 1600/1601-series (service & transactional) caller-ID rules, and BFSI/government sector requirements are all encoded, commercial calls to +91 must originate on an India route with a compliant CLI.
Three enforcement modes per org: off, shadow, and enforce. Shadow evaluates and logs every dial decision while blocking nothing, a safe rollout that produces audit evidence, then enforce blocks non-compliant calls before they're placed.
Explicit DLT-verified consent (OTP on the DLT Consent Registrar), inferred, and service-context tiers, each with evidence tracking. TCCCPR Feb-2025 rules are enforced: 90-day re-consent cooldowns after opt-out and 7-day verifiable-inquiry windows under the Third Amendment.
One suppression list across voice, WhatsApp, and email, a customer replying 'stop' on WhatsApp suppresses them from calls too. Spoken opt-outs are detected on call transcripts in English and Hindi. And it's fail-closed: if suppression status can't be verified, the call isn't placed.
Every dial decision is recorded with its rule evaluation, reasons, and consent evidence, an append-only trail designed to survive an enterprise or regulatory audit.
Entity registration and declaration tracking are built in, with a readiness checklist endpoint, so when enforcement tightens, your paperwork is already in the system.
Can't hand customer data to a vendor? Don't. Register your own database and every byte of business data, customers, conversations, memory, campaign records, routes there. Crawlii operates the AI; you keep the records. Auditors point at your database, not ours.
DLT registration itself and final legal sign-off remain the customer's responsibility, no vendor can automate that. What Crawlii automates is everything around it: the rules, the evidence, and the record of every decision.