NourishIQ is a full-stack nutrition app built on Next.js, Amazon DynamoDB, and Vercel. Instead of forcing users to log height and age, it infers nutritional targets dynamically from meal patterns — less friction, higher retention.
Why DynamoDB
Every read in NourishIQ follows the same pattern: fetch one user's data by date range. Single-table design with composite keys handles this in one query, under 5ms, at any scale.
PK: USER#
SK: MEAL##
TTL on expiresAt auto-expires Travel Mode and Plate Scanner records. No cron jobs. Two GSIs handle cross-entity queries for scan history and regional travel data.
Why Vercel
Edge deployment kept DynamoDB latency low with zero infrastructure config. Environment variables wired in two minutes. Deploy on push.
Key features
Smart 19-day meal calendar (5 past, 14 future)
Plate Scanner — snap any meal, get macros + healthier swap
Green Speaker Pill — hands-free voice cooking instructions
Travel Mode — regional dietary guidance with auto-expiring data
Apothecary — preservative-free wellness remedies
Monetization: Free / $12 Pro / $20 Premium with family plans.
DynamoDB on-demand means zero cost at zero users, predictable cost at a million. That's the stack I'd use in production — so that's the stack I built on from day one.
Top comments (0)