AIML - Senior Embedded Machine Learning Engineer- Edge ML

We are looking for a highly motivated and experienced Senior Embedded ML Engineer to join our team focused on enabling cutting-edge machine learning capabilities on resource-constrained edge devices. In this role, you will be at the forefront of innovation as we bridg the gap between state-of-the-art ML models and the realities of low-power, real-time, embedded hardware.

Description

You will play a key role in designing, implementing, and optimizing ML solutions for highly constrained compute environments. This is a cross-disciplinary role that blends expertise in embedded systems, computer architecture, and machine learning to unlock new applications in areas such as IoT, wearables, robotics, and autonomous systems. RESPONSIBILITIES: - Design and implement embedded ML pipelines on microcontrollers and custom SoCs with tight compute, memory, and power constraints. - Optimize and quantize deep learning models for real-time inference on edge platforms. - Develop and maintain low-level firmware in C/C++ to integrate ML models with custom hardware accelerators and sensors. - Conduct performance benchmarking, memory profiling, and bottleneck analysis across various embedded platforms. - Collaborate closely with ML researchers, hardware architects, and product engineers to co-design efficient ML solutions from model training to deployment. - Evaluate new edge ML techniques, compilers (e.g., TVM, TFLite Micro, CMSIS-NN), and toolchains to advance the team's capabilities. - Contribute to the overall system architecture with a deep understanding of embedded compute, memory hierarchies, and data flow optimization.

• Strong proficiency in C/C++ and Python, with a solid foundation in embedded firmware development.
• Deep understanding of computer architecture, particularly ARM Cortex-M/A cores, SIMD, caches, memory alignment, and DMA usage.
• Proficiency in model deployment tools and compilers such as TensorFlow Lite for Microcontrollers, TVM, ONNX Runtime, and custom model conversion pipelines.
• Demonstrated expertise in performance analysis, using tools like perf, valgrind, gprof, or hardware-specific profilers.
• Experience working with hardware interfaces such as SPI, I2C, UART, and integrating with sensors or custom accelerators.
• Bachelor's, Master's, or PhD or equivalent experience in Computer Science or a related field.

• Hands-on experience with deep learning concepts, including model architectures (CNNs, RNNs, Transformers), training workflows, and post-training optimization (quantization, pruning, distillation).
• Familiarity with embedded RTOSes (e.g., FreeRTOS, Zephyr) and real-time application constraints.
• Comfort with debugging low-level issues across software and hardware boundaries.
• Excellent problem-solving and analytical skills with a thorough approach.
• Most importantly: a strong curiosity, willingness to dive deep into unfamiliar problems, and an eagerness to learn and grow in a fast-evolving field.