I am an AI/ML Engineer passionate about building intelligent systems that bridge technology and defense.
As the Team Lead for Project TerraDefender, I specialize in designing deep learning architectures for terrain analysis, infrastructure mapping, and satellite intelligence systems.

I have lead and mentored as:

• ML Lead — DefenseTech Architects Team
• Secretary & ML Lead — REC IEEE CS Society
• AI/ML Lead — Intellexa Club

I explore how AI and machine learning can be leveraged to address complex problems across industries — from data understanding to intelligent automation and creative innovation.

AI-Powered Military IPB System
TerraDefender is a deep learning-based platform engineered for terrain intelligence, building extraction, and environmental threat analysis using satellite and aerial imagery.
It integrates terrain classification, infrastructure mapping, and geospatial AI to assist in defense planning and autonomous rover operations.

Core Modules:

• Terrain Analysis using CNN-based segmentation
• Building & Vegetation Extraction with U-Net
• Rover Integration for real-time decision support
• PDF-based Tactical Reports and GeoAI Overlays

Real-Time Wireframe Camera System
Project WraithCast captures and renders 3D skeletal wireframes of environments and objects using high-speed imaging and depth-sensing technology.
Designed for simulation, surveillance, and AR-based terrain visualization, it enables real-time 3D mapping and tactical visual reconstruction.

Highlights:

• Real-time wireframe rendering pipeline
• Depth and motion-aware 3D skeletal extraction
• Applications in AR, defense visualization, and reconnaissance

Context-Aware Hybrid AI System
The RAG Project combines information retrieval with generative AI to produce contextually rich and reliable responses.
It enhances the precision of LLM outputs by grounding them in factual sources — ideal for enterprise Q&A, knowledge systems, and intelligent assistants.

Core Features:

• Retrieval module with vector-based semantic search
• Integration with Transformer-based LLMs
• Context synthesis and answer generation

AI for Emotional Support & Wellbeing
An empathetic conversational agent that leverages NLP and emotion detection to provide supportive, private, and human-like mental health assistance.
It understands user intent, detects sentiment, and offers tailored coping strategies or self-help guidance.

Features:

• Sentiment & intent recognition
• Emotion-aware conversation flow
• Privacy-preserving response system

Real-Time Voice Transcription Tool
A PyQt5-based application that converts live or recorded speech into text with high transcription accuracy.
It provides waveform visualization and playback control, making it an ideal solution for note-taking, transcription, and voice documentation.

Technical Stack:*

• PyQt5 GUI
• Real-time speech recognition (Google Speech API / Whisper)
• Waveform visualization using Matplotlib

• Deep Learning & Model Optimization
• Multimodal and Generative AI
• Self-Supervised Learning
• Explainable and Responsible AI
• Human–AI Collaboration

— Ginni Rometty, Former CEO, IBM