DIP Diet RAG Agent
Developed an AI-driven Retrieval-Augmented Generation system for personalized nutrition guidance using IBM Watson Studio for model orchestration and deployment.
Timeline
1 month
Role
Fullstack
Status
CompletedTechnology Stack
Key Challenges
- Progress based memory management
- Genuine Data Sources
- Lower Accuracy when onboarding multiple users at once
Key Learnings
- Performance Optimization
- Follow-up chat using LangChain
- Multilingual Chat and Voice system
DIP Diet Agent: Your one and only platform for healthy life
Overview
Dip Diet Agent is an AI-powered wellness assistant built on IBM Cloud and watsonx.ai, designed to deliver personalized, context-aware diet recommendations. By combining scientific nutrition data, Retrieval-Augmented Generation (RAG), and intelligent reasoning, it empowers users to make informed health and dietary choices with ease.
Why We Built This
We created Dip Diet Agent to solve a real-world challenge — the abundance of generic, unreliable diet advice online.
While exploring nutrition and AI systems, we observed that most chat-based assistants lacked factual grounding, context awareness, and personalization.
Our goal was to build a trusted, science-backed AI diet companion that merges verified nutritional data with IBM Cloud’s scalable infrastructure and watsonx.ai’s powerful models, enabling users to achieve sustainable wellness through intelligent guidance.
Tech Stack
- IBM Cloud – Reliable, scalable infrastructure supporting AI workflows
- watsonx.ai & Granite-3.2-8B-Instruct – Contextual reasoning and human-like response generation
- Langflow – Modular RAG pipeline for efficient query handling
- AstraDB – Semantic vector storage for rapid content retrieval
- Verified Nutrition Data Sources – DIP PDFs, research papers, and dietary guidelines
Future Plans
- Convert this to mobile application for Android, iOS