Portfolio

• 

  RAG-based Compliance Intelligence System

  Code Repository: github.com/divya28jain/RAG-based-Compliance-Intelligence-System-with-FastAPI-Backend

  Built a Retrieval-Augmented Generation (RAG) system using FastAPI, FAISS and TinyLlama to answer compliance-related questions from PDF documents.

  Processed over 1100 text chunks from a 160+ page regulatory document and implemented semantic search for accurate retrieval.

  Added rule-based compliance risk detection for penalties, deadlines, and violations.

  Optimized local inference to achieve an average response time of approximately 22 seconds.

• 

  AI Travel Agent

  Live Project: ai-travel-agent-pro.streamlit.app/
  Code Repository: github.com/divya28jain/AI-Travel-Agent

  Built an AI-powered travel assistant that creates personalized itineraries based on destination, budget, and user preferences.

  Implemented a multi-agent workflow using Agno with Gemini API and SerpAPI for contextual travel planning.

  Developed a Streamlit interface with secure API key management and deployment.

• 

  Custom Object Detection Model

  Code: Google Colab

  Built a fruit detection system using Faster R-CNN.

  Applied image preprocessing, augmentation and hyperparameter tuning to improve detection accuracy.

  Evaluated the model using precision, recall and mAP metrics.

• 

  Dare Tiles - Augmented Reality Mobile Game

  Developed an Augmented Reality game using Unity and Vuforia.

  Implemented multiple image targets with randomized dare generation.

  Built gameplay using Unity physics and C# scripting.

• 

  Maze Runner

  Built a 2D maze racing game using Unity.

  Designed multiple levels with collision detection, physics-based movement and obstacles.

  Developed gameplay logic using C#.

• 

  VR Room Customization and Object Recognition Assistant

  Designed a VR application that recognizes room objects and allows users to customize virtual interiors.

  Built with Unity and C# while integrating object recognition concepts.