Ashok Kumar Kannan

Background

About Ashok Kumar Kannan

As a Technical Lead & Project Manager specializing in AI acceleration and compiler frameworks, I build systems that make AI faster, lighter, and more scalable. I currently lead a portfolio of projects focused on optimizing deep learning workloads for a custom AI accelerator - bridging the gap between software, hardware, and real-world AI performance.My expertise spans designing graph compilers for PyTorch, ONNX, JAX, and TensorFlow; building quantization and pruning frameworks for efficient inference on edge devices; and deploying high-throughput LLM pipelines using technologies like vLLM and Ray Serve. I also work at the intersection of Generative AI and Networking, bringing cutting-edge AI applications to new domains.I'm passionate about building systems that scale from compiler internals to real-time inference and mentoring teams to push the boundaries of what AI systems can do. My journey has been recognized through a triple promotion to Technical Lead in 2025 in a single year, the prestigious President’s Club Award at Multicoreware, and national-level accolades in AI and hardware innovation.I'm always eager to connect with teams working on disruptive AI hardware/software, ML systems engineering, or compiler-level innovation. If you’re building the future of AI infrastructure, I’d love to talk.E-Mail : ashokkumar.datascience@gmail.comMobile :+91 6382040562