Bangladesh Health Context
The product is tuned around Bangladeshi food, household budgets, family health decisions, and nearby care guidance.
Project Documentation
Understand NutriShastho AI from product idea to deployment.
This public documentation explains what the project does, how users interact with it, how the services are connected, which AI features are used, and how the system can be deployed.
Judge / Reviewer Notice: Free Tier Deployment
The backend and MCP server are deployed on Render's free tier. They spin down after 15 minutes of inactivity. When you first open the app, it may take up to 60 seconds for the servers to wake up. Please be patient on the first request!
Budget-Aware Nutrition
Diet plans consider monthly budget, family size, local food items, market area, and health constraints.
AI With Fallbacks
Provider-backed generation is optional. Local rules keep important workflows usable when AI calls fail.
Database-Backed Product
Neon/Postgres stores core data, while scheduled cleanup keeps generated history from growing forever.
Browse The Details
Documentation Sections
Each page is written for outsiders, judges, collaborators, and early users who need to understand the project without signing in.
User Journey
How To Use
A practical walkthrough from account creation to health input, budget planning, AI explanations, and care navigation.
Read sectionSystem Design
Architecture
NutriShastho AI uses a Next.js frontend, FastAPI backend, Neon/Postgres database, and an MCP server for AI tool orchestration.
Read sectionResponsible Intelligence
AI Features
The project combines deterministic safety rules, provider-backed generation, and MCP tools so AI output stays explainable and resilient.
Read sectionProduct Surface
Core Features
The dashboard brings together health tracking, nutrition planning, nearby care, reports, family context, and AI assistance.
Read sectionTrust Model
Data And Safety
The project uses Postgres for core records, explicit retention cleanup for generated history, and safety-first wording for medical output.
Read sectionFree-First Operations
Deployment
The project is prepared for Vercel, Render, Railway, Hugging Face Spaces, Docker Compose, and GitHub Actions.
Read section