Ratnesh Kumar Rai — AI Researcher

I’m a Senior AI Software Engineer with 5+ years of experience building NPU‑accelerated AI systems for edge and Android platforms, specializing in AI model optimization, inference frameworks, and production deployment using Intel AI technologies. I develop and optimize AI and GenAI models (including Gemini) for Intel NPUs and hardware IPs, focusing on quantization, performance tuning, and accuracy validation for Android workloads. I contribute to end‑to‑end NPU software enablement in the Intel Android AI stack, including neural network operations for the Android Neural Network HAL (NNHAL). I extend Intel’s inference ecosystem by implementing and integrating OpenVINO backends within the LiteRT framework, enabling low‑latency, high‑performance inference on resource‑constrained Android and edge devices. I also design NPU‑accelerated applications and reusable inference frameworks using OpenVINO and gRPC, supporting real‑world use cases such as object detection, face detection, LLM‑based chatbots, and Driver Monitoring Systems (DMS). My background includes Computer Vision, NLP, and Reinforcement learning, Generative AI and I’m passionate about building scalable, performant AI systems at the intersection of models, hardware acceleration, and production software.

Stackforce AI infers this person is a Senior AI Engineer specializing in AI acceleration and optimization for edge devices.

Location: Bengaluru, Karnataka, India

Experience: 4 yrs 9 mos