Case study · Nutrition and health
Custom Meal Planner AI Chat BOT.
A custom AI-powered chatbot platform built to generate personalized, GLP-1-friendly meal plans using a proprietary nutrition methodology.
- What we built
- Subscription AI chatbot for personalized meal plans
- Stack
- Next.js, TypeScript, Nodemailer, Supabase
- Services
- AI chatbotWeb appSubscription access control
About the project
What is Custom Meal Planner AI Chat BOT?
This project focused on the design and development of a custom AI-powered chatbot platform built to generate personalized, GLP-1-friendly meal plans using a proprietary nutrition methodology. The platform was conceived as a subscription-based SaaS product, combining advanced AI logic, structured nutrition rules, and a user management system with payment validation. The solution targets individuals using GLP-1 medications (such as Ozempic, Wegovy, and Mounjaro), as well as existing customers of a digital nutrition guide and new direct subscribers.
What it delivers
The chatbot delivers highly personalized meal plans, grocery lists, and ongoing dietary support while maintaining strict adherence to user-defined calorie, protein, and portion targets. At its core, the platform replaces static meal guides and manual planning with an interactive, intelligent, and scalable AI system. It is designed to support thousands of users simultaneously while maintaining accuracy, consistency, and a supportive, dietitian-led tone.
Conclusion
The chatbot transforms structured nutrition methodology into an intelligent, scalable SaaS platform. By combining strict nutrition rules, adaptive appetite logic, and subscription-aware access control, the system delivers a personalized and trustworthy experience for GLP-1 users. For the client, the platform represents a commercially viable, market-ready AI product that enhances customer value, drives recurring revenue, and scales without manual intervention.
The problem
What was slowing them down.
- 01
Manual effort and friction
Users had to calculate portions, calories, and grocery lists themselves, increasing drop-off and inconsistency.
- 02
Static meal guides lacked personalization
Existing guides required manual interpretation and could not adapt to appetite changes, preferences, or restrictions.
- 03
No intelligent cycling for appetite fluctuations
GLP-1 users often experience varying appetite levels, which traditional meal plans fail to accommodate safely.
- 04
No centralized AI solution
Generic AI tools lacked domain-specific rules and often produced inaccurate or unsafe recommendations.
- 05
Complex access rules tied to purchases
The business required differentiated access for existing customers, trial users, and paying subscribers, which could not be managed manually at scale.
- 06
What the client needed
The client needed a custom AI solution that combined nutrition intelligence, scalable access control, and a seamless user experience, without compromising accuracy or business logic.
The solution
What we built.
Rule-driven AI meal planning
We designed and developed a custom AI-powered chatbot platform that delivers personalized, rule-driven meal plans while integrating subscription logic and user management directly into the system. The chatbot collects structured inputs, applies strict nutrition rules, intelligently adapts to appetite patterns, and generates complete meal plans and grocery lists in a conversational format. Access is governed by real-time payment and subscription validation through external commerce APIs.
- Personalized AI meal planning: meal plans are generated based entirely on user-provided calorie, protein, and portion targets, without estimation.
- Structured nutrition framework: all meals follow a consistent methodology covering protein, produce, carbohydrates, and fats.
- Appetite-aware cycling logic: the system intelligently adjusts calories and protein across high- and low-appetite days while maintaining safety thresholds.
- Subscription-aware access control: user access dynamically adjusts based on purchase history, trial status, and active subscriptions.
- Scalable SaaS architecture: designed to support thousands of users with minimal operational overhead.
Results
What changed.
Users and the business
Highly personalized nutrition plans
Highly personalized nutrition plans tailored to GLP-1 usage.
Reduced friction
Reduced friction through automated planning and grocery lists.
Improved adherence
Improved user adherence via appetite-aware calorie cycling.
Scalable monetization model
Scalable monetization model with trials and subscriptions.
Lower support overhead
Lower support overhead through automated access control.
Market-ready SaaS product
Market-ready SaaS product positioned for growth. For the business, the chatbot transformed a static digital guide into an interactive, subscription-driven product.
Key challenges
The hard parts, and how we solved them.
Appetite variability on GLP-1 medications
Implemented a calorie and protein cycling system with configurable variance levels and safety constraints.
Preventing inaccurate nutrition estimates
Enforced strict use of user-provided values only, with no AI guessing or extrapolation.
Managing complex access rules
Built a user management system tied directly to purchase, trial, and subscription states.
Dietary restriction handling
Designed a rules-based exclusion and substitution system that automatically maintains nutritional balance.
User trust and tone
Adopted a supportive, educational, dietitian-style voice grounded in empathy and clarity.
Tech stack
What it runs on.
- Frontend & backend
- Next.jsTypeScriptNodemailer
- Database
- Supabase
- Knowledge base
- AWS Lambda
Process
How we got there.
01User Authentication & Access Validation
Secure, email-based magic link login · Real-time access check on each login, based on user type · Existing customers with permanent access · Trial users within a 14-day window · Active subscribers · Inactive or expired users are automatically blocked with clear messaging
02Onboarding & Input Collection
Plan duration (3, 5, or 7 days) · Portion-based or calorie-based tracking · Imperial or metric measurement preferences · Daily calorie and protein targets · Meal frequency and appetite patterns · Dietary restrictions, dislikes, and cooking preferences
03Meal Plan Generation
All days generated at once in a clean, readable text format · Nutrition targets per day · Appetite cycling rules · User preferences and restrictions · Desired meal repetition or variety
04Grocery List Creation
Consolidated grocery list, automatically organized by category with exact quantities matching measurement preferences
05Ongoing Interaction
Meal swaps · Image uploads (such as menus) · New chats while maintaining access history
More work.
All projects
EBAS - GPT
Multi-agent AI platform automates recruitment workflows, generating consistent documents, reducing turnaround time, and improving scalability with structured knowledge and exports.
Strategy Shortcut System Chatbot
Built a multi-agent AI SaaS platform helping businesses create strategies, execution plans, and marketing workflows through guided chatbot interactions.
Building something like Custom Meal Planner AI Chat BOT?