Debasmit Roy@debasmit2020

IssueCopilot.AI utilizes advanced AI to automate ticket generation by parsing customer reviews and directing them to appropriate developer teams. Using Mistral LLM models from TogetherAI, it extracts technical issues from sources like Play Store reviews and Reddit, storing data in Vectara for analysis. The RAG-powered chat interface fosters efficient team communication and decision-making. Teams benefit from visual data representation, identifying trends and prioritizing resolutions. Tickets are intelligently routed using the Llama index query pipeline based on issue scope and team expertise, ensuring prompt resolutions and iterative enhancements.

This project leverages Falcon LLMs for compliance, risk management, and real-time monitoring of LLM chatbots. Implementation Azure CosmosDB: Stores user info and violations. Multithreading in Flask: Enhances scalability. LRU Cache: Reduces latency for similar queries. Falcon LLM-powered Keyword Extractor: Creates a Rule Knowledge Graph (KG) from admin-uploaded PDFs, serving as an alternative to fine-tuning. Azure Blob Storage & Neo4j: Stores admin’s PDFs and the Knowledge Graph. Semantic Router: Retrieves relevant subgraphs and identifies violations. User Authentication: Clerk integrated with CosmosDB. Deployment: Frontend on Vercel, backend on PythonAnywhere. Objective Periodically updated Rules Knowledge Graph. Display rule violations, reasons, and priority levels. Support for multi-modal (image+text) query checks. Assign Risk scores based on violations. Analyze user behavior patterns. Methodology Flask Server: Multithreaded for handling requests and database queries. AI Pipeline: Rule Registration: Extracts and decomposes rules from PDFs, forms a knowledge graph, integrates external sources, and allows admin modifications. Query Processing: Degenerates user queries, retrieves relevant subgraphs, checks for rule violations, and reasons for violations. Risk Scoring Risk scores based on EU AI Act categories: Critical, High, Medium, Minimal. Analytics Admin panel displays rule violation frequency, custom timeframe violations, and user-wise analytics using Chart.js. Scope Compliance Assurance: Adherence to policies and regulations. Risk Management: Identifies and mitigates risks. Real-time Monitoring: Ensures transparency and accountability.

Reminisce.AI is an comprehensive, multi-modal, AI application designed to enhance the lives of individuals with Dementia, Alzheimer's, different types of Amnesia, providing personalized memory support, query resolution, and routine management. Built to run efficiently on edge devices, the app leverages Llama 3.2-1B-Instruct for offline processing and Llama 3.2-11B-Vision for vision-driven tasks, allowing it to function both online and offline depending on connectivity. This is complemented by speech-to-text, text-to-speech models for smooth, natural language, audio-based interaction. By integrating state-of-the-art SPLADE and CLIP embedding models, our app can understand both visual and textual data in a highly efficient manner, enabling responsive and context-aware interactions. The app uses a Knowledge Graph-based Retrieval-Augmented Generation (RAG) architecture, which allows it to store and retrieve detailed, personalized information, ensuring that users receive relevant and accurate responses based on their past interactions. With these powerful AI models, Reminisce.AI helps users relive cherished memories through image slideshows, offers real-time assistance with everyday queries, and provides scheduled alerts for meals, medication, and sleep routines, ultimately improving both the quality of life for individuals with dementia and peace of mind for their caregivers.