Software Architect

The AI Model / Algorithm Engineer focuses on developing and applying artificial intelligence and machine learning models to solve complex problems. This role designs, experiments with, and optimizes algorithms using methods from statistics, optimization, probability theory, and machine learning, and works with platform teams to deploy models into production.

Experience: 10+ years

Audio Signal Processing

• Design and implementation of audio processing algorithms including feature extraction, denoising, AEC, localization, beamforming
• Development and optimization for embedded deployment

Model development

• Develop ML based solution for audio use cases such as keyword spotting, noise suppression, audio classification and sound event detection: Training and optimizing deep learning models for such tasks
• Define performance metrics for ML model and audio use case
• Set up and maintain Linux-based development environments
• Configure and troubleshoot GPU training infrastructure
• Manage CUDA, cuDNN, Python environments and deep learning toolchains.
• Develop automation scripts for training and deployment workflows
• Build, validate, maintain audio datasets for model training

Model deployment on embedded systems

• Quantization, awareness of memory, compute and power constraints
• Analyze accuracy versus performance trade-offs

• Deploy ML models using TensorFlow Lite Micro (TFLM)
• Integrate inference pipelines on embedded DSP platforms
• Optimize inference performance on Cadence HiFi DSP processors
• Profile and debug performance bottlenecks

More information about NXP in India...

#LI-29f4