Qualcomm Technologies, Inc. is looking for an experienced individual for their on-device applied ML team to optimize ML models (GenAI, Vision AI, LLMs, VLAs) to on-device AI accelerators for optimal power, performance, and accuracy, creating edge-AI solutions for various applications.
Requirements
- Hands on Experience with ML model optimization and analysis
- Hands-on low-level programming experience with DSP, AI accelerator.
- Hands-on knowledge of TensorFlow, TFLite, Pytorch, ONNX
- Strong understanding of machine learning frameworks, tools, and technologies, with expertise in at least one major platform
Responsibilities
- Analyze and optimize various ML models on various classes on Hexagon NPU.
- Understand benefit/drawback of various quantization schemes, use internal tool chains to control loading, execution and scheduling of various models.
- Perform system level analysis of concurrent ML model execution and daisy chaining ML models including Vision AI, Gen AI and VLA models.
- Leverages expert Camera Engineering knowledge to research, design, develop, verify, debug, implement, and/or validate highly complex camera systems, HW or FW tasks (including Camera Image Signal Processor, 3A, Image Quality, Features, and Tuning Tools), algorithms, features, logic design, modeling, design verification, and power optimization.
- Performs architecture, IQ modules, algorithm, and feature decomposition to drive highly complex requirements and specifications for hardware development.
- Ensures advanced image quality, system performance, and highly critical ISP area and power metrics are met.
- Acts as a technical lead and facilitates collaboration across organizations to design, develop, and optimize camera systems, hardware, and/or software.
Other
- This leadership role will be hybrid in San Diego, CA, Santa Clara, CA and remote within the U.S.
- Work with world renowned customers from various fields like robotics, at home devices, drones, IP cameras, enterprise solutions and understand their ML needs.
- Provide technical leadership to R&D team
- Coordination between cross functional/geo teams
- Translates customer needs and requirements into long-term tasks and plans.
