Inspiration
I faced this problem myself as a product manager - spending weeks building features only to realize critical user flows were missing from our PRDs. Developers would constantly ask "what happens when..." and we'd scramble to figure it out mid-sprint.
What it does
Flowmender is an AI-powered PRD analyzer that: Validates if your document is actually a PRD (not a recipe or random doc!) Extracts 3-5 core user journeys from your requirements Identifies critical gaps, ambiguities, and missing business logic Generates structured analysis reports with actionable recommendations Saves analysis history for team reference Provides 3 free credits to get started + downloadable PDF reports Tracks usage with Microsoft Clarity analytics Deployed on Netlify with custom domain
How we built it
AI Engine: Google Gemini 2.0 Flash API for intelligent document analysis Backend: Supabase for user management, credit tracking, and analysis storage Frontend: React with modern UI components Analytics: Microsoft Clarity for user behavior insights Deployment: Netlify with custom domain integration
Challenges we ran into
System Prompt Engineering: Spent weeks fine-tuning prompts to get focused, actionable insights instead of generic business advice Gemini API Integration: Wrestling with response parsing and error handling for consistent JSON outputs Balancing Depth vs Speed: Making analysis comprehensive yet fast enough for good UX
Accomplishments that we're proud of
Real Users: Already have product managers(5-10) using it and providing feedback Actionable Output: Users report finding actual gaps they missed in their PRDs End-to-End Product: From idea to deployed product with payments and analytics
What we learned
You can build sophisticated AI tools without deep coding knowledge using modern APIs System prompt engineering is an art, small changes create dramatically different outputs Real user feedback beats theoretical features every time
What's next for Flowmender
User Research: Deep dive into PM pain points beyond just PRD analysis Monetization: Credit-based pricing model for sustained usage Team Features: Collaborative analysis and review workflows
Built With
- 2.0
- ai
- analytics:
- api:
- auth
- authentication:
- clarity
- database:
- deployment:
- flash
- framework:
- gemini
- javascript
- languages:-typescript
- netlify
- postgresql)
- react
- supabase
Log in or sign up for Devpost to join the conversation.