Machine Learning Engineer

They are currently seeking a Machine Learning Researcher with specialist skills in Audio Speech Recognition to take on a senior role in their small and friendly team.

You would help to set the research goals for their speech recognition project, with a particular focus on training the transducer ASR system. With the opportunity to work on all elements of the ASR system, your responsibilities would include the design and execution of experiments targeted at making algorithmic improvements, reducing the WER and latency, as well as optimising the inference results. You could see your ideas realised as hardware solutions and you will also support the team with production scale training. Opportunities to represent the company to clients are a possibility, as you will be working on the company’s key machine learning products. Excellent software and communication skills are a must, along with a driven and methodical approach to research in a commercial setting.

You will need:

• Masters or PhD level qualification in a related technical or scientific area with excellent understanding of machine learning and deep learning.
• An impressive commercial background including: scoping, planning and delivering ASR research projects; product focussed R&D and /or software delivery.
• Excellent skills in PyTorch, TensorFlow or JAX
• Full rights to work in the UK without time limit.

It’d be great if you also have:

• Good skills with C++ / CUDA or similar high-performance languages
• Good knowledge and understanding of model compression methods, such as quantisation, pruning and distillation.

An excellent remuneration package is available for the right candidate and the company enjoys a nice balance of working from home (2 days a week) and coming together as a team in the office (3 days a week).

Key words: Machine Learning, Audio, Automatic Speech Recognition, Cambridge, Senior, Hybrid-working, Research, Transducer ASR, Algorithm Optimisation, C++, CUDA, PyTorch, TensorFlow, JAX.