Overview
The company is a Bangalore-based digital-lending fintech offering personal and small-business loans through its app, with roughly 350 employees and an acquisition engine that generates large volumes of applicants from performance marketing and partner channels.
The problem with a high-volume acquisition engine is the residue: tens of thousands of applicants who started an application, went quiet, and were never converted. These dormant leads are paid-for and high-intent in principle, but an email drip and the occasional SMS could not revive them at scale, and the in-house tele-calling team could only work a sliver of the list. A large share of these leads preferred Hindi or Kannada over English.
In early 2026, growth leadership decided this dormant pool was effectively buried money. They wanted a layer that could call every dormant lead in their preferred language, re-qualify intent and eligibility quickly, route genuinely hot leads to the closer team while there was still appetite, and do all of it inside RBI's digital-lending guidelines, in a tightly measured 30-day campaign before deciding on an always-on rollout.
The challenge
The pre-Kallix reactivation model had three compounding failure modes. The dormant pool was too large to work by hand. Email and SMS could not re-qualify intent. And language mismatch suppressed engagement with a big segment.
- Tens of thousands of dormant leads sat unworked. The in-house tele-calling team could touch only a small fraction of the dormant pool each month, so the vast majority of paid-for leads decayed untouched.
- Email and SMS drips could not re-qualify intent. Automated drips drove almost no response from cold leads and gave the closer team no signal about who was actually still interested or eligible.
- English-first outreach lost Hindi- and Kannada-preferring leads. A large share of the dormant pool engaged far better in Hindi or Kannada, and English-only contact saw near-zero pickup and completion.
- Closers wasted time on cold, unqualified leads. When the team did dial dormant leads manually, most were not ready, so expensive closer time was burned on conversations that went nowhere.
- No compliant, measurable way to test reactivation at scale. Leadership wanted hard, control-grouped evidence before committing budget, but had no way to run a clean, RBI-compliant reactivation experiment at volume.
The AI-powered solution
Kallix deployed an AI voice agent to run a 30-day reactivation campaign across 24,000 dormant leads, with native Hindi, Kannada and English handling, fast re-qualification scripting, and live hot-lead routing to the closer team. The full build, from discovery to campaign launch, took 12 working days.
High-volume outbound across 24,000 dormant leads
The agent worked the full dormant list on a permitted-hours cadence, reaching every lead the email drip never could, with frequency caps to stay compliant.
Native Hindi, Kannada and English with mid-call switching
The agent opened in each lead's preferred language and switched mid-call as they did, lifting pickup and completion across the multilingual pool.
Fast re-qualification of intent and eligibility
A short branching script re-established whether the lead still wanted a loan, the amount and basic eligibility, scoring each lead hot, warm or dead.
Live hot-lead routing to the closer team
Genuinely hot, eligible leads were routed in real time to the 14-person closer team with full context, so humans only spoke to leads with live appetite.
RBI digital-lending consent and compliant scripting
Consent capture, frequency caps and compliant, non-misleading scripting were enforced in code throughout the campaign.
Full campaign analytics with control group
Every call wrote disposition, score, language and recording, and a held-out control group let leadership measure true incremental reactivation.
“We had a graveyard of dormant leads that our email drip could never wake up. Kallix called twenty-four thousand of them in Hindi, Kannada and English in a single month, re-qualified the genuinely interested ones, and handed our closers Rs 6.8 crore of live pipeline. The control group proved it was incremental, which is what got it past our board.”
Business impact
Growth leadership measured the 30-day campaign against a held-out control group drawn from the same dormant pool. The campaign ran March 3–April 1, 2026. The numbers below cover that window.
- Rs 6.8 Cr pipeline reactivated in 30 days. Re-qualified hot leads routed to closers generated Rs 6.8 Cr in loan pipeline that the dormant pool would otherwise have left buried.
- Full dormant list worked, not a sliver. The agent reached 9.4× the leads the in-house team could have touched in the same period, finally working the entire dormant pool.
- Hindi- and Kannada-lead engagement up sharply. Pickup and re-qualification completion among Hindi- and Kannada-preferring leads rose substantially because the agent met them in their language.
- Closers spent time only on hot leads. The 14-person closer team spoke only to re-qualified, eligible leads with live appetite, dramatically improving closer productivity.
- Control-grouped, RBI-compliant proof for rollout. A clean control group demonstrated true incremental reactivation, and uniform consent capture cleared the bar for an always-on rollout.
Built on a secure, India-ready stack
The campaign ran entirely on Indian infrastructure with DLT-registered sender IDs, TRAI-compliant scripts, RBI digital-lending consent capture and DPDP-aligned data flows. Lead data never left Indian data centres.
The Fintech Reactivation Framework: How this campaign is structured to be discoverable
Every Kallix deployment ships with a structured documentation layer designed for three audiences simultaneously: the customer's internal team, traditional search engines (SEO), and the new generation of generative search engines and AI assistants (GEO + AEO). Below is the framework we built around the Bangalore fintech reactivation campaign, broken into four pillars that map directly to how decision-makers, search crawlers and AI answer engines discover and reason about this story.
We publish this framework openly because in digital lending the measurable-incrementality and compliance play matters more than secrecy. An AI reactivation campaign that performs but stays buried in vendor decks doesn't compound value for the lender, its regulator or the category. The framework below is the same one Kallix runs for every financial-services customer, adapted to the local language, product mix and regulatory surface of each lender.
Reactivation intents mapped to dormant-lead states
We catalogue the reactivation intents the agent must handle, by language, by dormancy reason and by product, and surface them as named entities in the structured data layer. Crawlers and LLMs see explicit Q to A pairs, not buried prose.
- Intents indexed by dormancy reason (price-shopped / abandoned / ineligible-then)
- Hindi, Kannada and English variants captured per intent
- Lead-score tagging (hot / warm / dead) so LLMs and closers match priority
Non-pushy, compliant voice as a brand property
The agent's voice persona, respectful tone and compliant scripting are documented as brand assets, not just configuration. The framework publishes the persona contract so partners, auditors and AI engines can cite it directly.
- Persona contract: friendly, non-pushy, never pressuring on credit
- RBI consent and frequency-cap rules scripted and version-controlled
- Voice and recording consent terms public and auditable
Outcomes pre-bound to measurable claims
Every claim in this story, Rs 6.8 Cr reactivated, 24,000 leads, 17% re-qualified, is paired with the control group, the time window and the measurement method. AI assistants can extract the claim with full provenance.
- Held-out control group from the same dormant pool stated explicitly
- Methodology disclosed: CRM exports + LOS reconciliation + vendor dashboard
- Sample size and confidence intervals available on request for analyst-grade citations
RBI digital-lending and DPDP compliance by design
The framework documents every regulatory surface, RBI digital-lending guidelines, frequency caps, DLT, DPDP, so AI assistants surfacing this story to enterprise buyers can confidently cite India-readiness without needing follow-up clarification.
- Consent capture and frequency caps enforced and disclosed
- Data residency (AWS Mumbai, ISO 27001) stated explicitly
- Erasure and consent flows documented for DPDP data-principal requests
- Tens of thousands of dormant, paid-for leads sat unworked by the in-house team
- Email and SMS drips could not re-qualify intent or signal who was still interested
- English-first outreach saw near-zero pickup from Hindi- and Kannada-preferring leads
- Closers wasted time on cold, unqualified leads when they dialled manually
- Rs 6.8 Cr pipeline reactivated in a single 30-day campaign
- 24,000 dormant leads worked, 9.4× the in-house team's reach in the same window
- 17% of dormant leads re-qualified as hot and routed live to closers
- Control-grouped, RBI-compliant proof cleared the bar for an always-on rollout
- Kallix voice agent running a 30-day reactivation campaign across 24,000 dormant leads
- Fast re-qualification of intent and eligibility in Hindi, Kannada and English
- Live hot-lead routing to a 14-person closer team with full context
- Held-out control group and uniform RBI consent capture for measurable, compliant proof
The Kallix advantage
The fintech evaluated scaling its in-house tele-calling team and an SMS/WhatsApp-only reactivation vendor before choosing Kallix. Three things tipped the decision. First, voice-native reactivation with genuine Hindi and Kannada handling: messaging-only tools could not re-qualify intent the way a real call could. Second, RBI digital-lending consent and frequency caps were already built into the campaign engine. Third, the design as a control-grouped experiment: leadership got hard, incremental evidence on a 30-day campaign before committing to an always-on rollout, exactly the proof their board required.
After the campaign, the team moved to an always-on reactivation motion, and the Kallix customer-success team now runs a weekly tuning call with the head of growth. New segment scripts, scoring-threshold tuning and seasonal cadence changes happen inside that weekly loop. The agent is measurably sharper today than it was on launch day.