AI Engineer specializing in Generative AI, Computer Vision, and full-stack systems. I translate research into production solutions using deep learning, multimodal frameworks, and robust engineering practices.
Currently pursuing Computer Science (AI/ML) at VIT Chennai. Ex-Intern at Larsen and Toubro. Delivered solutions across healthcare AI, Software Defined Vehicles (SDVs), compiler optimization, and financial intelligence.
🏆 3rd Place - Synthetic Data & OpenEnv Challenge (October 2025)
MAROONED — Multi-agent deception environment with LLM-guided reinforcement learning. Recognized for novel approach to emergent behavioral dynamics and teacher-guided learning architecture.
Sponsored by: 
• PyTorch • AMD
MAROONED: Multi-Agent Deception Environment
Custom OpenEnv implementation with 5 simultaneous agents, asymmetric information dynamics, and teacher-guided SFT. 10,000-step episodes with sparse rewards testing long-horizon language reasoning and emergent deception strategies.
NeuroGPT: AI-Powered MRI Insights
Multimodal AI framework for neuroimaging diagnostics. 99.68% specificity, 70% reduction in documentation time.
Generative Medical Imaging Pipeline
GAN-based MRI modality translation with Grad-CAM interpretability. ~88% SSIM. Under review (Elsevier, IF 2.1).
OptiMap: AI Compiler for pPIM
Compiler for 24-bit Processor-in-Memory architectures with AI-guided optimization.
Financial Intelligence Platform
Market analysis pipeline with transformer-based sentiment analysis. 60% reduction in manual workflows.
Unreal FPS AI Prototype
UE5-based first-person shooter prototype. Implements player controls, weapon mechanics, and Blueprint-based AI with patrol, detection, and attack behavior. Built as a sandbox to explore gameplay logic and AI systems in Unreal Engine.
AI/ML: PyTorch, TensorFlow, Transformers, OpenCV, Scikit-learn
Backend: FastAPI, Flask, gRPC
Cloud: AWS, GCP, Docker, Kubernetes
Frontend: React, Next.js, Flutter
Data: MySQL, Firebase, DynamoDB