Adi Polak
November 16, 2022

Our community is full of people with incredible skills and know-how. And this nomination proves us right!

Our community member @Leonard Aukea has been nominated for Machine Learning Professional of the year as part of the Nordic DAIR Awards. Congratulations, Leonard! 

Who is Leonard?

Leonard Aukea has been Heading Machine Learning Engineering and Operations at Volvo Cars for the last 4.5 years. As part of his journey, transforming Volvo cars machine learning platform and taking it to the next level, he  joined lakeFS community and  has been proactively shaping the open-source product, while sharing his learnings from building Volvo’s Machine Learning platforms. He also spoke about his journey in multiple public occasion and recently on a podcast discussing how structuring Volvo for operational success has took the platform capabilities for new highs. His recognition is definitely well-deserved.

At Volvo Cars, Leonard is responsible for developing the overall mission and strategy for ML Engineering and Operations and leading the process of building reproducible ML systems. 

Why vote for Leonard

During the 18 month we have been collaborating with Leonard we had the privilege to work with an innovative, forward thinker, who is not only an exceptional research professional, but also understands what it takes to scale an ML operation and stream line in to bring real business value. Leonard knew the pieces that would make a great MLOps infra for Volvo, and he went out of his way to either bring in or create those. His commitment to open source, and willingness to contribute from his learnings and experiments to the products he is using had been a model for us all in the lakeFS community

More on his innovation and leadership:

Furthermore, under his leadership, Volvo Cars developed a git-centric approach to ML experimentation and delivery, adopting MLOps principles through cultural transformation and tooling that opens the door to efficient development throughout the ML lifecycle. lakeFS is part of Volvo’s ecosystem of tools and infrastructure, helping engineers run ML experiments in fully isolated development environments and streamline their ML platform. 

During his many public appearance , Leonard shared that the ML teams work with hundreds of terabytes (TBs) of sensor data generated by engines and image/video data from the cars. All the data is stored in Amazon S3 and on prem in different formats:

  • structured IoT data in tabular format (some in Parquet file format), 
  • unstructured videos and images data encoded in a proprietary format, 
  • and other data sets mix structured and unstructured data.

Some of the stored video files are as much as ~80 TB, requiring teams to manage data effectively without duplication.

Volvo adopted lakeFS across its machine learning platform and different file formats to drive Data Lifecycle Management utilizing lakeFS isolated branching capabilities and Data version control. This, together with Spark, AWS S3, Kubeflow and Tecton Feature Store, makes this platform complete.

Volvo Machine Learning Architecture – Reference: https://www.youtube.com/watch?v=VzgomadGo1g

Volvo data versioning – Reference: https://www.youtube.com/watch?v=VzgomadGo1g

About the Nordic DAIR Awards

Nordic DAIR Awards recognize great work in data, analytics, and AI across the Nordics, celebrating exceptional individuals, teams, and organizations that foster talent, make great contributions to the field, and drive data and AI Innovation forward.

The Machine Learning Professional of the year award is granted to practitioners that show outstanding leadership or expertise in ML – developing new algorithms, systems, and AI-based solutions that bring proven value and impact to the business and organization.

We’re thrilled to see a member of our community get recognized that way. We encourage you to vote for Leonard by following this link: https://hyperight.com/dair-awards/individual-categories-2022/ 

LakeFS

  • Get Started
    Get Started
  • Git for Data - What, How and Why Now?

    Read the article
    +