Ph3Experteer Overview /h3pAs Staff AI/ML Embedded ML/DSP Systems Engineer, you will lead the architecture and optimization of real-time audio AI systems across industrial, data center, and wearables domains. You work at the hardware-software frontier, shaping DSP/NPU deployment, and guiding model compression and fixed-point implementations. You’ll collaborate with RTL and ASIC teams to ensure hardware-aware algorithm design and robust validation. The role offers mentorship, strategic impact on AI/ML architectures, and involvement in cutting-edge audio processing. This is a chance to contribute to Analog Devices’ mission at the Intelligent Edge and advance Physical AI initiatives. /ph3Retribuzione / Benefits /h3ulliArchitect and optimize end-to-end deployment pipelines for compact audio AI models on DSP/NPU targets /liliDefine DSP/NPU partitioning strategies balancing workload, memory, latency, and power across the SoC /liliOwn simulation-to-RTL validation flows with bit-exact reference models and RTL co-simulation /liliImplement and optimize fixed-point signal processing and neural network kernels for efficient inference /liliProfile and optimize inference performance under always-on, real-time constraints for hearables/wearables /liliDesign and maintain model compression/quantization workflows (PTQ, QAT) with quality tracking /liliDevelop array processing algorithms (beamforming, spatial filtering) from prototype to fixed-point deployment /liliContribute to audio ASIC system architecture decisions based on algorithmic and deployment needs /liliGenerate IP and represent technical depth to OEM customers in automotive and hearable segments /liliMentor engineers in deployment practices and hardware-aware algorithm design /li /ulh3Responsabilità /h3ulliMasters/PhD in Electrical Engineering, signal processing, or related field /lili6+ years in audio/speech signal processing within semiconductor environments /liliHands‑on deployment experience on DSP and/or NPU platforms /liliExpertise in fixed‑point algorithm implementation and model quantization (PTQ/QAT) /liliStrong knowledge of simulation-to-RTL flows and bit‑exact modeling /liliProficiency in C (embedded/firmware), Python, MATLAB, and deep learning frameworks (TensorFlow/TFLite, PyTorch/ONNX) /liliExperience with low‑level profiling tools, ISA, and memory optimization for embedded AI /li /ulh3Requisiti fondamentali /h3ullicompetitive compensation and benefits /liliwork-life balance /liliopportunity to work on cutting‑edge projects /li /ul /p #J-18808-Ljbffr