Hello, I'm Anup Bhat

• Developed an AI-driven platform to improve financial literacy across diverse age groups in India.
• Implemented a Contract Analyzer using NLP to detect harmful clauses in financial documents.
• Built a multilingual voice assistant using BharatGPT and AI4Bharat to support users in native languages.
• Gamified financial education with simulation tools mimicking real-life financial decisions.
• Added a Financial Knowledge Radar to detect learning gaps and recommend content.
• Used YouTube Data API and Google News RSS for personalized and curated educational content.

2024 – Present

• Developed an educational app that gamifies learning experiences, starting with mathematics, featuring a level map with various topics and point-based reward system (+10 points per correct answer).
• Implemented a unique biomimicry section encouraging students to observe nature and find innovative solutions inspired by millions of years of evolution.
• Created engaging PvP quiz systems within the biomimicry section where students compete directly, fostering healthy competition and collaborative learning.
• Built interactive map features providing insights into architecture inspired by nature, combined with augmented reality (AR) showcasing biomimicry structures.
• Integrated Firebase for real-time data synchronization, user authentication, leaderboard management, and social features for knowledge sharing.
• Developed WebAR features using Blender for 3D modeling and visualization of biomimicry structures, encouraging students to brainstorm innovative solutions.
• Designed scalable architecture with future scope for AI-powered personalized learning, gamified teacher assessment tools, and global competitions.

• Developed a domain-specific chatbot for financial queries using MeTTa knowledge graphs and Graph RAG for intelligent retrieval.
• Implemented a hybrid response system combining structured knowledge graphs with Google Gemini as a fallback LLM for complex queries.
• Created a custom Python interface for MeTTa integration and Streamlit-based web interface for user interaction.
• Built a React.js frontend with an interactive UI for seamless user experience and real-time query processing.
• Designed a scalable architecture supporting multiple financial domains including loans, credit scores, and debt management.
• Integrated advanced features like legal transparency checks and financial literacy guidance through the knowledge graph system.

May 2025 – Present

• Developed a comprehensive nutrition app aimed at improving Indian dietary habits through heritage-focused meal planning and AI-powered recipe optimization.
• Implemented "Restore My Recipe" feature using AI to modify traditional family recipes while maintaining nutritional heritage and cultural authenticity.
• Built intelligent recipe discovery system with personalized filters for dietary preferences, regional cuisines, budget constraints, and health goals (diabetes, PCOS, muscle building).
• Created comprehensive food logging system with forward-focused nutritional advice that encourages better eating habits rather than dwelling on past choices.
• Designed dual optimization tracks combining modern nutritional science with traditional ancestral wisdom to maximize health benefits.
• Integrated Gemini AI and Groq LLM for generating concise, actionable nutritional recommendations targeting Indian families and Gen-Z users.

2024 – Present

• Built a comprehensive web application for analyzing student performance data from competitive exams (JEE/NEET) with AI-powered feedback generation.
• Implemented detailed performance analytics including overall metrics, subject-wise breakdowns, chapter-level insights, and time management analysis.
• Integrated multiple LLM support (Gemini and Groq) for generating personalized coaching insights and strategic improvement recommendations.
• Created interactive visualization suite with donut charts, bar charts, and time analysis plots using Matplotlib for comprehensive data representation.
• Developed professional PDF report generation system using ReportLab for exportable performance analysis with embedded charts and AI feedback.
• Designed scalable JSON data processing pipeline supporting various competitive exam formats with real-time analysis capabilities.

May 2024 – Present

• Built an AI-powered app that generates personalized interview questions based on resumes to ease interview anxiety.
• Used Natural Language Processing (NLP) to extract key details and craft relevant technical and behavioral questions.
• Implemented an intuitive interface using Streamlit, integrating LangChain for advanced prompt engineering.

2025

• AI-powered multilingual assistant for police stations that uses a Neo4j knowledge graph to provide legal guidance and complaint support across kiosk, web, and mobile platforms.
• Integrates Google Gemini for conversational AI, Sarvam AI for voice support in 12 languages including Konkani, and Redis for context management.
• Offline JSON caching ensures uninterrupted service.
• Supports voice complaint filing with smart form autofill, while a police dashboard tracks cases and communications to streamline operations.

2025

• A speed-reading platform designed for users with ADHD to improve focus using an auto-scrolling reader synced to their WPM.
• Features a distraction-free UI with customizable fonts and themes, plus a built-in reading speed test.
• Supports EPUB uploads, personal library management, and progress tracking.