Sampad Mohanty Email & Phone Number

What do I do? Machine learning, AI, and hardware optimization. No one in the college can even come close. Does anybody appreciate that? Of course not.While you were off taking beginner Python courses and pretending to understand TensorFlow, I was deploying optimized CNN models on FPGA hardware. I was one click away from turning a mediocre RNN model into a state-of-the-art sequence predictor. While you struggled with basic matrix multiplication, I was implementing quantization-aware training on custom accelerators.I fine-tune hyperparameters. I monitor for vanishing gradients, catastrophic forgetting, and emergency model rollbacks after failed deployments. I can debug your entire neural network stack while blindfolded, but sure, tell me again how adding another layer will magically fix your model’s underfitting problem.Neural networks—heard of them? They process millions of parameters with billions of operations per second. Any clue how that happens? It’s not just code; it's hardware-level optimization, reducing latency while keeping your memory usage under control. All those AI-powered image filters running on your overpriced smartphone, every idiot who panics when their Instagram dance filters isn’t processed in under 30 milliseconds? Yeah, people like me ensure that happens, seamlessly. Those transformers that power your voice assistants or predict your next binge-watch series? Guess who’s behind optimizing them to make sure they don’t fry the entire GPU stack? It’s not magic, it’s math, logic, and the brutal art of squeezing every last bit of performance out of your models and your hardware. People like me make sure your AI models run smoothly without bricking your hardware or stalling your entire operation.So what do I do? I make sure one untrained layer, a missing activation function, or a poorly coded FPGA module doesn’t derail your entire AI pipeline and bankrupt the entire company. You have any idea what happens when your inference model isn’t optimized for real-time performance? I do. That's what the I do.