Case study · Education Technology

Voice AI for EdTech: Multilingual Outbound That Remembers.

EdTech outbound calling in India hits a wall that generic dialers cannot solve. Students and parents expect conversations in their own language, and a single script does not work across career guidance, parent-teacher follow-up, enrollment, feedback, and re-engagement. Kathan Voice OS handles all of these natively across 12+ Indian languages on the Alchemyst context layer, which keeps a persistent memory of every lead. Retargeted cohorts consistently outperform cold outreach because context accumulates across attempts.

38.7%Aggregate Connection Rate

Context, Not Cold Scripts.

01 · The problem

Multilingual Outbound at Scale Breaks Generic Dialers

EdTech operations in India span career guidance for new prospects, parent-teacher follow-ups for existing students, enrollment for upcoming batches, feedback collection, and re-engagement of leads who went cold. Each motion has its own opening, tone, and decision logic, and most run across at least four languages. Traditional manual dialing requires a large, expensive team. Generic AI dialers achieve 20 to 25% connection rates and treat every retry like a cold call, so retargeted leads convert worse than fresh ones, the inverse of what should happen when the institute already has prior signal on the lead. The signal exists. Nothing carries it into the next call.

02 · The context layer

Voice Agents That Carry Context Across Attempts

Kathan does not work from a flat script. The context layer gives each call a live, filtered view of the lead's history, the campaign's objective, the language preference, and the prior interaction trail, and writes the outcome back when the call ends. A retargeted Gujarati parent receiving a third PTM follow-up call hears a conversation that references the prior objection, the specific student, and the upcoming event by name, in Gujarati from the first syllable. The agent calling a Telugu CA student about exam prep operates from a different context entirely.

  • Persistent context across attempts: retargeted leads are not called from zero, and connection compounds across waves.
  • Per-campaign context scoping so career guidance, PTM, enrollment, feedback, and retargeting each follow their own conversational logic.
  • Six to twelve languages handled natively per deployment, with regional phrasing rather than English scripts dubbed through TTS.
  • TRAI-compliant caller ID, time-window enforcement, and complete audit trails.
03 · Where it fits

Where It Fits in the Student Lifecycle

Five workflow patterns repeat across coaching institutes, online platforms, and certification academies. Each one absorbs counsellor or admin bandwidth that should be focused on the high-intent conversations only humans can have.

  1. 01Career Guidance Outreach

    first-touch conversations with new prospects in their preferred language, qualifying interest before a counsellor ever joins.

  2. 02Parent-Teacher Follow-Ups

    reminders and rescheduling for PTM events, with parent-language conversations that respect regional norms.

  3. 03Course Enrollment Drives

    outbound calls during open windows that reference the specific course and prior interactions.

  4. 04Feedback & NPS Collection

    post-batch and post-event feedback at multiples of email response rates.

  5. 05Lapsed-Lead Re-Engagement

    retargeting waves that consistently outperform cold outreach because the agent already knows the lead's prior objection.

04 · Why it works

Why It Works for EdTech Specifically

Education is a relationship business, and the relationship is rarely English-only. A Gujarati parent expects a Gujarati conversation, and the call is lost in the first ten seconds otherwise. The compounding effect of memory across attempts is also more visible in EdTech than almost any other vertical: by the third touch, a retargeted cohort can connect at 1.5x the cold-outreach rate because every prior interaction sharpens what the agent says next. The result is an outbound motion that actually scales without scaling the headcount behind it.

More case studies

Same context layer, a different industry

Each of these connects different systems and runs different workflows. The memory underneath works the same way. All case studies.

Real Estate
7Products Unified

Real Estate Services

Unify the Stack. Enrich the Agents Already in Place.

Read the case study
Auto Retail
6Workflows Per Dealership

Automotive Retail

Every Reminder, Every Renewal, Every Adviser Smarter Than the Last.

Read the case study
HR Services
6Workflows Across the Talent Lifecycle

HR Services & Staffing

From First Inbound Call to 90-Day Onboarding, Without Losing Context.

Read the case study
Talk to us

Every conversation should start from what you already know.

Tell us which systems hold your customer's history and which workflows keep starting from zero. We will show you where the context layer slots in, without replacing what already works.

Talk to usStart building free