Problem
Indian SMEs (50-500 employees) — manufacturing units, BPOs, retail chains — face 25-40% annual attrition but rarely run structured exit interviews because HR teams are 1-2 people stretched thin. Without exit data, leadership only spots root causes (pay, manager issues, commute) after attrition becomes a crisis, costing ~₹30-50K per replacement hire in recruitment and training.
Solution
Build a voice AI agent (using telephony + LLM infra like Vapi) that calls departing employees within 48 hours of resignation, conducts a 10-minute structured exit interview in Hindi, English, or regional languages, and auto-generates a summary with sentiment tags. HR gets a dashboard aggregating attrition reasons by department, manager, and location, with monthly trend alerts. v1: outbound calling, transcription, LLM summarization, and a simple web dashboard.
Why Now
Voice agent infrastructure (Vapi, trending on Product Hunt in June 2026) has made it possible for small teams to build reliable, low-latency multilingual voice AI agents without owning a telephony stack. Combined with India's SME hiring boom and rising attrition costs, HR teams are actively seeking low-cost automation — Tracxn's June 2026 India data shows strong SME-tooling demand concentrated in Tier 2/3 cities.
Target User
HR managers at 50-500 employee companies in Tier 1-2 Indian cities (Pune, Coimbatore, Indore, Jaipur) — manufacturing units, BPOs/call centers, retail chains, logistics — with 10+ monthly exits, already paying for HR software (Keka, Darwinbox, Zoho People).
Business Model
SaaS subscription priced ₹4,999-14,999/month by employee count, sold as an add-on via existing HRMS marketplaces. Voice AI cost per interview is ~₹15-20, easily covered at this pricing, yielding ~75% gross margin at scale. Distribution via Keka and Zoho People app marketplaces.
Competitive Landscape
- Direct (India): none focused specifically on voice-led exit-interview automation; HRMS players (Darwinbox, Keka) offer generic text-based exit workflows
- Direct (global reference): Vapi (voice agent infra, Product Hunt June 2026), ExitPro/Voiceform-style exit-interview SaaS in the US
- Why we win: Multilingual (Hindi/Tamil/regional) voice interviews built for shop-floor and BPO workforces, plus integration-first distribution through existing HRMS marketplaces
6-Month Plan
- Month 1-2: Build voice agent (telephony + LLM) for Hindi/English exit interviews, transcript + summary pipeline (~₹3L)
- Month 3: Build HR dashboard (attrition trends, sentiment by department), pilot with 3 SMEs (~₹2L)
- Month 4: Add Tamil and Marathi, refine sentiment/topic tagging (~₹1.5L)
- Month 5: Integrate with Zoho People / Keka APIs, onboard 10 paying customers (~₹1.5L)
- Month 6: Marketplace sales push, target 25 paying customers, hire 1 sales/CS person (~₹1L)
Risks
- Adoption risk: HR teams may hesitate to let an AI call departing employees (mitigate with opt-in, HR-branded calls)
- Distribution risk: scale depends on negotiating HRMS marketplace partnerships, which take time
- Data sensitivity: exit interview data requires careful DPDP Act-compliant storage and access controls
Score Breakdown
- Market Size (13/20): SME HR-tech is a growing mid-sized segment — TAM among Tier 1-2 SME HRMS users is roughly ₹300-500Cr, not yet ₹1000Cr+
- Capital Efficiency (13/15): MVP achievable for ~₹3-4L using off-the-shelf voice AI and LLM APIs; 12-month runway fits comfortably under ₹20L
- Team Feasibility (8/10): 2 developers plus 1 HR-domain advisor can ship v1 within 8 weeks
- Trend Momentum (12/15): Voice agent infra (Vapi) trending on Product Hunt in June 2026, plus India SME-tooling demand signal from Tracxn
- Moat & Defensibility (11/15): Aggregated, anonymized attrition benchmark data across customers builds a data network effect over time
- Unit Economics (12/15): SaaS subscription with ~75% gross margin once voice AI costs are amortized
- Time-to-MVP (7/10): Core voice agent and dashboard achievable in ~8 weeks; HRMS integrations add a few more weeks