Nithin Sai Kumar Kopparapu — AI Researcher & Deep Learning Engineer
From vision-language models to edge deployment — turning research into real-world AI
Research, deployments, and experiments across the full AI stack.
Chain-of-Thought prompting and QLoRA fine-tuning to combat sycophancy in small VLMs. 78% safety improvement.
Sequential multi-task learning without catastrophic forgetting of previously learned knowledge.
UAV-based signaling system for road efficiency and collision prevention at intersections.
Real-time detection and counting on NVIDIA Jetson Nano & Xavier using optimized YOLOv4.
Classical CV algorithms, filters, and transformations in Jupyter Notebooks.
Comparative study of VAE and GAN architectures for synthetic training data augmentation.
Can small LMs learn arithmetic? Comparing supervised pretraining and RL fine-tuning paradigms.
DQN with experience replay and target networks solving the CartPole benchmark.
Immersive puzzle game with WebXR VR/AR support, 3D manipulation, and hint system.
Microcontroller programming, sensor integration, and real-time control in C.
Hardware prototyping with sensor integration and IoT experiments.
Image processing & CV algorithms for monitoring clogging in glass furnaces. Image segmentation using classical and ML-based approaches.
5G RAN deployment, Kubernetes cluster installation (RHOCP), O-RAN deployment and sanity testing.
5G RAN platform deployment, cluster installation, GNB testing.
What I work with, day in and day out.
Open to opportunities in AI, Deep Learning, and Computer Vision research across Europe.