Initiative Type: AI-Driven Consumer Tool (Expense Management)
Portfolio Status: Beta-ready case study for public portfolio (PDF-ready with visuals and workflow link). Try Splitwiser live app.
Phase 1: Discovery & Opportunity Analysis – Origins and Problem Space
Problem: Shared expense management for roommates and group settings was consistently unfair, error-prone, and a source of tension. Existing tools like Splitwise and Excel lacked the ability to handle partial participation and itemized breakdowns, forcing users to overpay or compromise, damaging trust.
Context: As a student living in shared housing, repeated frustrations over grocery, takeout, and utility bills highlighted that current tools weren't solving for the human need: transparency and trust, not just calculation.
Research: Conducted 20+ interviews with students and young professionals:
- 85% wanted automation and fairness, admitting to skipping tax/tip or rounding to avoid conflict.
- Most did not want another heavy app, but desired quick, chat-friendly solutions.
- Users trusted AI-calculated fairness over spreadsheets or manual math.
Opportunity: Build an AI-driven, chat-first tool that:
- Automatically parses real-world bills (grocery, utilities, dining).
- Allows item-level toggling for precise fairness.
- Integrates with familiar channels (WhatsApp, Telegram) to avoid adoption friction.
Phase 2: Strategy & Definition – Shaping the Solution
Vision: Make splitting group expenses instant, transparent, and conflict-free — acting as a neutral "roommate mediator" rather than another calculator.
Personas:
- Student Roommate – Needs fast, fair solutions, budget-driven.
- Young Professional – Time-sensitive, expects automation and minimal manual work.
- Travel Organizer – Manages group trips, wants one-off bill settlement without disputes.
Core MVP Features:
- GPT-powered bill parser that extracts line items, tax, and tips automatically.
- Per-person, per-item toggling for exact shares.
- Proportional tax and tip distribution.
- Exportable summaries for group chats (WhatsApp/Telegram) without requiring signups.
Competitive Positioning:
- Vs. Splitwise: Adds true item-level fairness, faster onboarding, no persistent account.
- Vs. Spreadsheets: Removes error and manual work; faster to settle.
- Vs. Venmo-style notes: Offers structured, auditable clarity with no ambiguity.
Phase 3: Execution – Bringing the MVP to Life
MVP Build: Developed a GPT-driven parser and a lightweight web-based interface capable of creating households, adding members, scanning bills, toggling participation, and generating share breakdowns.
Testing Milestones:
- Shared MVP with 20 early adopters (students and professionals) to gather structured feedback on accuracy, UI, and user trust.
- Next testing iteration will incorporate real-world bill formats (mixed grocery + takeout receipts, utilities) to refine parsing robustness and edge case handling.
Workflow Visual: A complete diagram illustrating how users upload receipts, toggle items, and receive a transparent split summary will be featured in the PDF (linked above).
Phase 4: Validation – Measuring Success
Metrics to Track:
- Adoption rate across early testers.
- Average time to finalize and settle a bill.
- Reduction in disputes compared to manual or Splitwise methods.
Feedback Channels:
- In-app satisfaction surveys.
- Direct user interviews.
- Retention tracking over recurring household cycles.
Iteration Priorities: Focus on features that increase trust (clear summaries, transparency) and reduce effort (fast parsing, zero-manual-entry flow).
Phase 5: Impact & Growth – Expanding Reach
Launch Strategy:
- Initial launch as a free-to-use web tool targeting university students and young professionals.
Marketing Initiatives:
- Share the beta in university WhatsApp groups to drive feedback and adoption.
- Place posters with QR codes in grocery stores and common spaces near university housing.
- Collect survey data to refine product positioning and adoption strategy.
Expansion Roadmap:
- Integrate with Telegram API to launch a Telegram bot, enabling one-click bill splitting directly in chat groups.
- Explore integrations with WhatsApp and Discord post-beta.
- Future monetization via premium features (bill history, analytics, API licensing for platforms).
Phase 6: Retrospective & Learnings – Post-Launch Goals
- Identify which features drive trust and adoption.
- Pinpoint drop-off stages in the workflow and optimize user flow.
- Document learnings on building and scaling AI-powered consumer tools for future iterations.