Data Engineer - Multimodal Systems

As a Data Engineer - Multimodal Systems, you will be a core contributor to creating, collecting, and improving Zyphra’s datasets and data pipelines across a variety of modalities. Your work will intersect with almost every team at Zyphra. You will be involved in collecting large-scale datasets and implementing and optimizing highly parallel data pipelines.

• Large-scale data collection across a variety of modalities (text, audio, image)

• Designing and working with highly efficient, parallelized data processing pipelines across modalities

• Designing and running rigorous experimental ablations to demonstrate the impact of new data improvements

• Strong implementation and prototyping ability

• Can take an idea from conception to experimentation quickly

• The ability to work well with others in a high-paced research setting

• Can rapidly learn new fields and are excited to implement new ideas

• Excellent communication and collaboration skills, and can work effectively on both research and engineering implementation at scale.

• Experience collecting, handling, and processing large datasets

• Experience with parallel Python programming frameworks such as Dask

• Understanding of the state-of-the-art in dataset curation across modalities

• A generally meticulous nature and a strong interest in actually looking at data and sanity checking things

• Strong grasp of proper experimental methodology for running rigorous ablations and other hypothesis testing

• Understanding of and interest in large-scale, highly parallel data processing pipelines.

• Proficiency with PyTorch and Python.

• Experience contributing to large pre-existing codebases and rapidly getting up to speed.

• Previously published machine learning research in well-respected venues.

• Postgraduate degree in a scientific subject (Computer Science, EE/EECS, Mathematics, Physics, Machine Learning)