Problem
India's MSMEs frequently need formal Detailed Project Reports (DPRs) and CMA (Credit Monitoring Arrangement) data to apply for Mudra, CGTMSE, PMEGP, and state subsidy-linked bank loans, but most first-time borrowers can't produce bank-grade financial projections themselves. CAs and loan consultants charge ₹3,000-15,000 per DPR and take 1-2 weeks, causing many small applicants to abandon formal credit applications or get rejected on documentation grounds alone.
Solution
A web app where an MSME owner answers a guided questionnaire (business type, machinery/working-capital needs, revenue plan, scheme chosen) in English or Hindi; an LLM pipeline trained on bank-approved DPR/CMA formats generates a complete project report - projected P&L, balance sheet, cash flow, ratio analysis, and scheme-specific annexures - as a downloadable PDF/Excel ready for bank submission. v1 covers Mudra and CGTMSE formats for 5 common MSME categories (retail, food processing, manufacturing, services, transport).
Why Now
Product Hunt's June 2026 weekly leaderboard featured "VC Boom," an AI tool that scores pitch decks and drafts investor-ready materials - showing AI-generated, finance-grade document automation is now trusted for high-stakes fundraising documents. The same LLM capability, applied to standardized bank loan templates, is newly viable for India's lending-document bottleneck.
Target User
First 1000 customers: first-time MSME loan applicants in Tier-2/3 towns (Coimbatore, Jaipur, Nashik) - shopkeepers, small manufacturers, and service operators seeking ₹2-50 lakh loans under Mudra/CGTMSE, who currently pay a local CA or loan agent and want a faster, cheaper alternative, reached via referral from bank DSAs and CA networks.
Business Model
Per-report fee of ₹999-1999 paid via UPI, or a ₹4,999/month plan for CA firms and loan consultants generating multiple DPRs for clients. COGS is mainly LLM inference (~₹30-50/report), giving over 90% gross margin on direct sales and ~75% margin on the CA subscription tier after support costs.
Competitive Landscape
- Direct (India): Local CAs and loan consultants (informal, offline); no dedicated AI DPR-generation SaaS yet
- Direct (global reference): VC Boom (US) - AI pitch-deck scoring and investor-document drafting that inspired the document-automation approach
- Why we win: Purpose-built bank-template library plus instant turnaround (minutes vs weeks) at a fraction of CA fees, distributed through CA firms and bank DSAs
6-Month Plan
- Month 1-2: Build questionnaire + LLM pipeline for Mudra DPR format, 2 MSME categories (~₹2.5L)
- Month 3: Add CGTMSE format + CMA data tables, pilot with 2 CA firms (~₹1.5L)
- Month 4: Expand to 5 MSME categories, add Hindi questionnaire (~₹1.5L)
- Month 5-6: Onboard 10 CA firms via subscription tier, add PMEGP format (~₹1.5L)
- Total spend ~₹7L of the ₹20L budget, leaving runway for marketing and support hires
Risks
- Banks may treat AI-generated DPRs as non-credible without a CA's signature, limiting acceptance (high likelihood, high impact)
- Scheme formats (Mudra/CGTMSE/PMEGP) change periodically, requiring ongoing template maintenance (medium likelihood, medium impact)
- CA firms may see the tool as a threat to fee income and resist adoption rather than use it as a productivity tool (medium likelihood, medium impact)
Score Breakdown
- Market (15/20): Tens of millions of MSMEs apply for formal credit annually, and DPR/CMA generation is a recurring need across every loan cycle - a large document-services TAM rather than a platform-scale one.
- Capital Efficiency (13/15): MVP is primarily an LLM pipeline plus templated PDF/Excel generation, achievable within ₹7L.
- Team Feasibility (8/10): 2-3 people (1 dev, 1 finance/CA domain expert for template accuracy) can ship v1 within 10 weeks.
- Trend Momentum (9/15): VC Boom's June 2026 Product Hunt traction validates AI-generated finance documents, but it's a US fundraising tool, not a direct India-lending analog.
- Moat (8/15): Bank-template library and CA-network distribution create some lock-in, but the core LLM capability is replicable by competitors.
- Unit Economics (12/15): Over 90% margin on per-report sales; LLM cost per report is minimal relative to price.
- Speed (7/10): Single-scheme, single-category MVP achievable in ~10 weeks using existing LLM APIs and template design.