AK Sinha — Software Engineer

I am an experienced ML systems engineer with over 5 years of expertise in delivering production-grade solutions in edge AI and accelerator-driven environments. My passion lies in designing end-to- end model execution frameworks that facilitate the seamless deployment of deep learning workloads on resource-constrained hardware. I specialize in developing compiler-centric optimization strategies that effectively bridge high-level ML abstractions with efficient low-level execution, ensuring optimal performance across various platforms. Throughout my career, I have demonstrated a strong ability to design and optimize production-grade inference engines for edge ML accelerators. My work has involved leveraging advanced graph compiler techniques to convert ONNX models into highly efficient C++ implementations. I am also knowledgeable in hardware-aware model transformation techniques, which ensure the optimal utilization of compute, memory, and dataflow architectures. My commitment to performance engineering has driven significant latency reduction and throughput improvements for real-time inference applications. 🏆 Key Achievements: 👉🏼 Honored with the AMD Spotlight Award (2025) for impactful contributions to the FlexML Graph Compiler and advancing the AI Engine-ML architecture for Ryzen AI NPUs and Versal adaptive SoCs. 👉🏼 Recognized with the AlphaICs Excellence Award (2022) for outstanding performance in graph compiler development, ML kernel optimization, and building scalable CI/CD infrastructure for the RAP™ architecture.

Stackforce AI infers this person is a specialist in AI hardware and compiler optimization for edge applications.

Location: Bengaluru, Karnataka, India

Experience: 4 yrs 2 mos