All work
04 · Health, AI recommendation systems
Alyve Wellness: AI recommendation system for personalized supplements from questionnaires and blood reports
- Client
- Health platform
- Industry
- Health, AI recommendation systems
- When
- 2023, ongoing
- My part
- Full-stack work alongside the core team
Overview
Alyve Wellness recommends supplements from a person's own questionnaire answers and blood reports, not from a demographic. An AI recommendation system turns that input into a personalized plan. I contributed across the front end, API and data layer.
What I built
- 01Questionnaire flows and blood report intake.
- 02The AI recommendation system and the DynamoDB data model behind it.
- 03React and Next.js front end, Node.js API on AWS.
What was hard
Problem
Health data has to be handled carefully.
Approach
Access controls on AWS and separation between identity and health records.
Problem
Query patterns on DynamoDB for a recommendation workload.
Approach
Designed keys and secondary indexes around the actual access patterns rather than a relational model.