Proof of Usefulness Report
FinPal
Analysis completed on 6/13/2026
+72.11
Proof of Usefulness Score
You're In Business
FinPal demonstrates high real-world utility and excellent technical innovation tailored to an underserved Indian UPI market. However, despite the massive TAM, the project is in a nascent prototype phase with minimal verifiable traction, no reported active users, and no revenue. The technical implementation, particularly the fine-tuned mBERT and LSTM layers, shows significant promise if user adoption can be accelerated.
View All Reports
Score Breakdown
Real World Utility0.25) × 1.0
Audience Reach Impact0.20) × 0.5
Technical Innovation0.15) × 1.5
Evidence Of Traction0.25) × 1.0
Market Timing Relevance0.10) × 1.0
Functional Completeness0.05) × 1.5
Subtotal+64.38
Usefulness Multiplierx1.12
Final Score+72
Project Details
Project URL
Description
I built FinPal, an AI-powered personal finance assistant for Indian UPI users. This isn't a demo. This is the raw, real story of what worked, what broke, and what the data actually told me about 350 million people's spending habits.
Audience Reach
India has 350M+ active UPI users — the single largest real-time payment network in the world. FinPal targets the 22–35 age bracket within this base: salaried professionals and gig workers who transact daily but have zero financial visibility. This is an underserved segment ignored by traditional banking apps. The total addressable audience is conservatively 80–100 million people in India alone, with expansion potential across any UPI-enabled market.
Target Users
Indian salaried professionals and gig economy workers aged 22–35 who use UPI daily (Swiggy, Zomato, Amazon, rent, groceries) but have no clear picture of where their money goes each month. They are not served by traditional personal finance apps built for Western markets. FinPal speaks their language — literally (Hinglish insights) — and works with the payment infrastructure they already use every day.
Technologies
Other, Python, HuggingFace Transformers (multilingual BERT), PostgreSQL + TimescaleDB, React Native, OpenAI API (GPT-4o-mini), FastAPI
Traction Evidence
GitHub repo with active commit history. The transaction categorizer achieves 91% accuracy on real Indian UPI SMS data after fine-tuning on 14,000 hand-labeled messages. ML inference runs on EC2 t3.medium at ₹1,200/month total infra cost for early users. Anomaly detection precision improved from 67% to 84% after adding LSTM temporal layer. Early user testing shows 34% reduction in churn after implementing confidence-threshold UI (auto-confirm high-confidence categories, prompt user on <80% confidence). Real users making spending decisions based on FinPal insights.
Algorithm Insights
Market Position
Growing utility with room for optimization
User Engagement
Documented reach suggests active user community
Technical Stack
Modern tech stack aligned with sponsor technologies
Recommendations to Increase Usefulness Score
Document User Growth
Provide specific metrics on user acquisition and retention rates
Showcase Revenue Model
Detail sustainable monetization strategy and current revenue streams
Expand Evidence Base
Include testimonials, case studies, and third-party validation
Technical Roadmap
Share development milestones and feature completion timeline