Crafting intelligent systems at the intersection of machine learning, deep learning, and human-centered design.
I'm an aspiring AI/ML Engineer with a strong foundation in machine learning and deep learning. I enjoy building industry-level technologies and transforming complex data into actionable insights.
My work spans neural networks, NLP, computer vision, and IoT systems — continuously pushing the boundaries of what's possible with AI.
A showcase of my work in AI, Machine Learning, and IoT systems.
Implemented a centralised file-sharing network with dynamic IP allocation using DHCP and secure FTP-based file transfer across multiple departmental subnets, connected via static routing and DHCP relay in Cisco Packet Tracer.
Built an intelligent prediction system using deep learning. Performed rigorous EDA, feature engineering (lags, rolling statistics, seasonality encoding) and time-aware train–test splitting, achieving improved forecasting accuracy measured via RMSE, MAE, R², and MAPE.
Built an intelligent conversational chatbot using transformer models trained on Cornell Movie Dialogs datasets, enabling human-like responses through contextual understanding and self-attention mechanisms. Fine-tuned the model for relevance, tone consistency, and response quality using techniques like beam search and sampling.
Developed a robust peer-to-peer file sharing application using WebRTC, designed to work seamlessly across mobile hotspots and various network topologies. Implemented reliable ICE candidate handling and connection fallback mechanisms to ensure stable and fast file transfers.
Developed a ChatGPT-like chatbot that runs completely on a local laptop using phi3, llama3, and mistral via Ollama. Works fully offline with no internet or API dependency, ensuring total data privacy. Built with Python and Streamlit, featuring custom system prompts and chat memory.
Developed Swallet, a personal finance tracking app using Streamlit, enabling users to log income and expenses in INR with intuitive UI and date selection via Google Calendar integration. Implemented data visualization using pie charts and graphs for spending insights; ensured local CSV-based storage to maintain user privacy.
Designed and implemented a microcontroller-based automatic room light system with a bidirectional counter using an infrared sensor and Arduino Uno, enabling lights to toggle based on human presence. Developed the control logic using Embedded C, integrated infrared relay switching, and documented the design with flowcharts.
Open to collaborations, research, and engineering roles in AI / ML.