Company Overview
Our client is a rapidly growing semiconductor innovator focused on developing ultra-low-power system-on-chip (SoC) solutions for intelligent, battery-powered devices worldwide. Their technology powers next-generation consumer, industrial, and IoT products, delivering advanced features while maintaining exceptional energy efficiency. With hundreds of millions of chips shipped globally, the company continues to push the boundaries of endpoint AI and sustainable computing.
Headquartered in the U.S., with additional teams in Asia and Europe, the organization fosters a collaborative culture centered on continuous innovation, technical curiosity, and practical problem-solving. They seek engineers who thrive in dynamic environments, enjoy tackling complex challenges, and are motivated by shaping the future of low-power AI computing.
This position is based in Austin, TX or San Diego, CA, requiring on-site presence five days per week.
Role Scope
The Endpoint AI team focuses on deploying machine learning (ML) and deep learning (DL) models across power-optimized SoCs. Using model compression and optimization techniques, the team brings sophisticated AI workloads to devices with tight compute, memory, and power constraints.
The ideal candidate is passionate about efficient computing and comfortable in a fast-paced, first-generation (“version zero”) development environment, leveraging internal, open-source, and third-party tools to deliver robust AI capabilities for real-world applications.
Key Responsibilities
* Design, refine, and deploy ML/DL models optimized for highly constrained embedded environments
* Apply pruning, quantization, and distillation methods to balance performance, accuracy, and power efficiency
* Develop, train, and document models for internal repositories and customer-facing AI libraries
* Collaborate with cross-functional hardware and software teams for optimal model integration
* Contribute to technical publications, conferences, and community engagements to share best practices
Required Skills & Experience
* Hands-on experience with model compression techniques for CNNs or RNNs
* Domain experience in audio classification, speech recognition, computer vision, or time-series analysis, including feature extraction
* Strong proficiency with TensorFlow (TFLite, TFLite Micro) and/or PyTorch
* Familiarity with dataset creation, curation, and augmentation for embedded ML tasks
Preferred Qualifications
* Background in TinyML or on-device AI
* Experience with microcontroller or embedded C/C++ environments
* Knowledge of attention-based architecture compression
* Familiarity with model-to-binary compilers (IREE, MicroTVM, etc.)
* Experience using ONNX, TOSA, JAX, LLVM, or MLIR
* Understanding of heterogeneous compute environments (CPU + NPU + DSP)
Education & Compensation
Bachelor’s degree in Computer Science, Electrical Engineering, or related field (Master’s or PhD preferred)
