EcoXpert is an AI-powered eco lifestyle agent designed to help individuals adopt sustainable living practices through personalized, real-time suggestions. By leveraging Retrieval-Augmented Generation (RAG) and IBM Watsonx services, EcoXpert empowers users to make greener choices through practical tips, policy guidance, and eco-friendly product insights.
In the age of climate change and environmental degradation, people are increasingly aware of the need to adopt eco-friendly practices—but lack of localized, actionable guidance makes it difficult to start. Most information is scattered across websites, policy documents, or hard-to-find sources.
EcoXpert bridges this gap by acting as a smart, conversational assistant. Powered by RAG (Retrieval-Augmented Generation), it draws insights from indexed documents to answer real-time queries such as:
- “How do I reduce plastic use at home?”
- “Which government schemes support solar energy?”
- “What are eco-friendly product alternatives?”
EcoXpert delivers instant, factual, and localized guidance tailored for Indian users, helping them adopt sustainable habits and leverage relevant government schemes.
- Watsonx.ai Studio
- Cloud Object Storage
- Vector Database & Indexing
- NLP (Natural Language Processing)
- RAG (Retrieval-Augmented Generation)
- IBM Cloud IAM
- IBM Granite LLMs
- Watsonx.ai Studio
- Watsonx.ai Runtime
- Watsonx Vector Index
- Cloud Object Storage (COS)
- Personalized sustainable living tips
- Localized waste segregation and recycling advice
- Government policy and subsidy recommendations
- Product category suggestions (non-branded)
- Support for natural language queries
- Environmentally-conscious citizens
- Urban and rural Indian households
- NGOs promoting sustainability
- Schools and eco-clubs
- Green-tech startups
- Localized Intelligence – Designed for Indian context, including region-specific data.
- Document-Grounded Responses – No hallucinations; every answer is traceable to a source.
- Simplicity & Usability – Beginner-friendly, jargon-free explanations.
- No Bidding, No Bias – No product or brand promotion.
- User Enters a Query – E.g., “What are eco-friendly travel options?”
- Agent Uses RAG – Retrieves relevant info from indexed PDF/Markdown docs.
- Granite Model Generates Response – Uses IBM LLMs for contextual response.
- User Gets Instant Guidance – Actionable, factual, and region-specific.
Agent Build Configuration
Configuration Settings
Knowledge Base Upload
Deployment Status Preview
Agent Testing Preview
-
Clone the repository:
https://github.com/Sarthaklad1034/EcoXpert-AI-Agent.git cd EcoXpert-AI-Agent -
Upload your knowledge base documents (PDF or Markdown) into IBM Watsonx.ai Vector Database.
-
Configure the RAG pipeline in Watsonx:
- Create a new agent
- Upload indexed documents
- Add Agent Instructions and Common Instructions
-
Test the assistant with natural queries.
-
Promote and deploy the agent through IBM Watsonx UI.
- IBM Watsonx.ai Documentation
- Granite Foundation Models
- India Government Sustainability Schemes
- Government Schemes for Sustainability in India
This project is licensed under the MIT License. See LICENSE file for details.
Project Lead: Sarthak Lad
Email: [email protected]
GitHub: github.com/Sarthaklad1034
“Small, consistent actions can build a greener future — let EcoXpert guide the way.”