Vani OS · Admissions Voice + WhatsApp

Turning the inquiries you already have
into twice the class.

Your 10,000 leads aren't a demand problem. They're a triage problem. About 75% of a calling floor's leads were never going to close — but nobody knew that going in. Vani OS is the AI admissions conversion layer that sits on top of the counsellor team or outsourced BPO floor you already have — fixing triage, speed-to-lead, context, and follow-up discipline on the calls you're already making.

001234567890001234567890%
of leads called today with zero qualification
001234567890001234567890×
more effective: a 5-minute response vs. 30 minutes
001234567890001234567890,001234567890001234567890001234567890
per incremental admit — payable only above baseline
001234567890001234567890%+
leaner closer team, same or better conversion
The bottleneck

Demand isn't the problem. Intelligent triage is.

Universities already generate real inquiry volume — 10,000 to 100,000 leads a cycle. Conversion leaks after the lead arrives, because volume is handed to callers without scoring, context, or an SLA clock.

HitListPanel

Bulk-imports inquiries from CSV or the CRM, scores each one against an ideal-student profile, and assigns prioritised hit lists — with an offline cache.

PitchAdvisorPanel

Contextual briefing before the first dial: course interest, prior touches, and likely objections.

FollowUpIntelPanel

Overdue, due-today, and warm-but-unclosed leads surface automatically — the SLA and next-best-action engine.

CrmApprovalPanel

The system prepares CRM updates; a counsellor confirms before anything writes back.

PipelinePanel

A relabelable stage pipeline — inquiry → applicant → admitted → enrolled — mapped onto your own program names.

TheFloorPanel

The daily counsellor rhythm: morning briefing → prioritised list → AI prep → live session → clean close.

Why this isn't "a CRM with a voice bot bolted on"

Six deterministic guardrail gates — not an LLM trying to be careful

Every admissions decision, legal/visa/medical topic, binding commitment, vulnerable-caller disclosure, minor-and-consent situation, or fabricated-urgency/discount attempt hits a phrase-matched hard gate before the agent can respond — checked, not hoped for. Rules are live-editable in the Counsellor Definition Studio, and every change requires a stated reason and is logged: who, when, why, before and after.

"You're not being asked to trust a black box. You're being shown the formula, watching the safety net catch a bad response live, and handed the receipt for every change anyone makes to it."
Not a roadmap slide

What's actually running today

The operator app for the admissions voice agent — six real, live screens, not mockups.

Console
Live

Live call queue — transcript, lifecycle stepper, extracted fields, listen-in audio.

Leads
Live

Admissions inquiries — the intake step ahead of a call or WhatsApp conversation.

Analytics
Live

Cost, outcomes, escalation categories, and the per-call ledger.

Call History
Live

Browse completed calls: outcome, cost, escalation category, filter and sort.

Counsellor Definition Studio
Live

Escalation rules, per-institution catalogue, and the state machine — as a UI, with a full audit trail.

Demo Guide
Live

Run of show, talking points, and a troubleshooting table for showing the live agent to someone new.

Benchmark validation

Sourced, not invented

85%
Faculty/interview reviewer headcount reduction — NMIMS Mumbai, 13,000+ AI-evaluated interviews
Dr. Sharad Mhaiskar, Pro Vice Chancellor
+100%
Enrollment growth — Sarala Birla University, Jharkhand (1,500+ students)
Counsellor conversion 5–10% → 15–20%
+30%
Admission efficiency from automated lead distribution and scoring alone
8,000+ student Indian institute
+25–35%
Conversion lift from WhatsApp automation in India
Industry benchmark
Commercial model

Priced on enrolled outcomes, not software seats

A hybrid structure: clear setup, transparent platform & ops, and the majority of the economics tied to incremental enrolled students above a locked baseline.

One-time setup
₹4–5L

CRM integration (LeadSquared / Zoho), stage mapping, SLA configuration, counsellor training, dashboards.

Platform & ops
₹3–4L/mo

Typically 4 months for one admission term — platform, AI briefings, the SLA engine, automation, and support.

Outcome fee
₹15,000/admit

Only on incremental enrolled students above a locked, same-period-last-year baseline. Nothing if we don't beat it.

Keep your CRM. Keep your team. Add the engine.

Pilot two programs, one admission term — live within 2–3 weeks, running alongside your existing team or BPO.