Research Engineer - Speech/Audio Machine Learning

On-siteMidResearch Scientistposted 10mo ago

✦Role summaryAI-generated

This role focuses on developing and training cutting‑edge foundation models for speech and audio applications in a research‑driven environment. You will collaborate with world‑class researchers and engineers to tackle fundamental challenges in AI and contribute to transformative industry solutions.

Skills required

Speech recognition Audio processing Deep learning Transformer models Python PyTorch Model training Research AI Machine learning Data pipelines Distributed training Docker Model serving Software engineering

About this role

About the Institute of Foundation Models

We are a dedicated research lab for building, understanding, using, and risk-managing foundation models. Our mandate is to advance research, nurture the next generation of AI builders, and drive transformative contributions to a knowledge-driven economy.

As part of our team, you’ll have the opportunity to work on the core of cutting-edge foundation model training, alongside world-class researchers, data scientists, and engineers, tackling the most fundamental and impactful challenges in AI development. You will participate in the development of groundbreaking AI solutions that have the potential to reshape entire industries. Strategic and innovative problem-solving skills will be instrumental in establishing MBZUAI as a global hub for high-performance computing in deep learning, driving impactful discoveries that inspire the next generation of AI pioneers.

The Role

As a Research Engineer specializing in speech/audio machine learning, you will bridge the gap between research and production constraints to deliver competitive performance. You will be responsible for the end-to-end performance of our machine learning speech systems, from architecting scalable training infrastructure to achieving extreme low-latency inference.

Key responsibilities

Compute Optimization: Optimize model compute graphs for target runtimes using frameworks such as TensorRT, Apache TVM, or OpenVINO.

Model Compression: Perform quantization (INT8/FP8) and pruning to enable real-time execution on edge or cloud GPUs.

Data Orchestration: Build reliable data pipelines for audio datasets, including automated denoising and synthetic data augmentation.

Distributed Training: Help design efficient multi-node training scripts using frameworks such as DeepSpeed, Horovod, as well as other distributed training technologies.

Deployment: Build and maintain high-performance demo environments on the cloud using Docker and standalone model serving frameworks. Integrate WebRTC or similar bidirectional protocols to ensure seamless, real-time audio-to-audio interactions.