Evidence-based nutrition education and coaching product
Atleet Buddy WhatsApp Nutrition Assistant
An interactive WhatsApp AI assistant for I'm a Foodie readers, designed to turn the Eet als een atleet book into a practical nutrition support experience with grounded answers, clear AI boundaries, credit-based usage, subscriptions, and admin oversight.
- LLMs
- React
- Supabase
- Dashboard UI
Readers needed practical, fast answers from book content while the team needed reliability, credits, subscriptions, and oversight
A more stable WhatsApp AI assistant experience grounded in book content with stronger product controls
Strengthened product readiness by improving assistant reliability, usage visibility, subscription awareness, and admin controls for a WhatsApp-first nutrition product.
The business friction
The product needed to transform a trusted book into an interactive assistant without becoming a vague health chatbot. Sports nutrition questions can be personal, nuanced, and sensitive, so the assistant needed to stay grounded in the source material, avoid medical advice, handle follow-up questions, maintain consistent language, support trial and paid usage, and give admins visibility into users, credits, subscriptions, usage, and estimated AI costs.
The system response
MASK AI improved the assistant behavior and the product layer around it. The work covered answer quality, retrieval context, fallback logic, refusal behavior, multi-turn conversation handling, and language consistency, plus the commercial infrastructure around trial access, prepaid credits, subscriptions, usage tracking, cost visibility, and admin controls.
What the system covers
Built around the operational jobs that matter: intake, product experience, workflow control, visibility, and follow-through.
WhatsApp-based AI assistant experience
Book-grounded answer behavior
Clear fallback and medical-safety boundaries
Multi-turn follow-up support
Trial, credit, and subscription visibility
User and plan management
Usage analytics and estimated AI cost tracking
Admin dashboard for product oversight
Admin controls and support visibility
Foundation for a commercially managed AI product
How the system flows
WhatsApp questions move through grounded context, safety checks, answers, credits, and admin review.
WhatsApp questions move through grounded context, safety checks, answers, credits, and admin review.
Stack and architecture
The engagement combined product interface work, workflow logic, and implementation tooling around the operational needs of the project.
- LLMs
- React
- Supabase
- Dashboard UI
- Usage Analytics
- Role-Based Access
What changed
The project helped Atleet Buddy operate as a more managed commercial AI product: clearer for readers, safer in scope, easier to monetize, and easier for the team to monitor as usage grows.
