Virtual Agent for Users to Help Them While Coding

---

Team Name: 8282 Squad

---

Team Members:

1. Omkar Sonawane
2. Rahul Gupta
3. Mohammad Alim
4. Siddharth Jain

---

Project Context:

The objective of this project is to develop a chatbot that enhances the learning experience by:

• Providing answers to frequently asked questions (FAQs) from each module to reduce the ticket count for repetitive doubts.
• Ensuring the chatbot is context-specific within the scope of each sprint.
• Guiding learners with hints rather than providing direct solutions.

Technologies Used:

Category	Technologies
Frontend	TypeScript, React.js, Context API, Axios, Tailwind CSS
Backend	Express.js, MongoDB, CORS, Node.js
LLMs and Backend	Flask, Langchain, Hugging Face Embeddings, FAISS Vector DB, GooglePalm (as LLM)

Implementation:

We implemented a Retrieval-Augmented Generation (RAG) based agent to handle FAQs for Qkart and the sales team. The model was deployed locally using Flask.

Usage:

• Access the frontend application via your browser.
• Interact with the chatbot to get FAQs related to Qkart and sales.
• The chatbot will provide hints and guide the learners through their questions without providing direct solutions.

Acknowledgments:

We thank the Crio.do team for organizing this hackathon and providing us with this learning opportunity.

Website UI Design

1. Sign up

2. Login

3. Chat History

4. Select Options

5. Milestone Selection

6. Conversations