Lightly secures $3m seed round to clean up data for machine learning

Machine learning is like software that is updated with data instead of code. Tech startup Lightly has taken on the mission to drive quality data into machine learning models to generate better outcomes.

Lightly’s data curation tool enables machine learning teams to save time and money on data selection and labelling, whilst at the same time improving the performance of models.

Matthias Heller, Co-founder of Lightly commented: “Most companies that use AI need to label data. But, the amount of data produced in a single day cannot be labelled in a million years. That's why it is so important to wisely choose what data to label and put into a machine learning model. That's where Lightly steps in.

"We deploy self-learning algorithms that tell data teams which one percent of their data is the most valuable. With our approach, companies see their labelling costs decrease by up to 90%, while their AI model can improve by 20%.”

Lightly was founded in 2020 by co-founders Matthias Heller and Igor Susmelj, former ETH and Harvard students, who worked in autonomous driving and research on computer vision and deep learning. They witnessed firsthand how many engineers' efforts are excessively directed into manually cleaning datasets or how data quality is disregarded and the model becomes the main focus.

The Swiss startup has taken an unprecedented approach to solving this data challenge. Their open-source framework with over 2k stars on GitHub has caught the attention of many industry leaders, such as Facebook’s Head of AI, Jerome Pesenti, who tweeted about it.

Lightly is a significant tool for a variety of reasons. It tackles the challenge of biased data, attempting to make AI more ethical, for instance. Additionally, it powers machine learning models with recent data ensuring they remain relevant. Lightly manages such data changes by intelligently adding novel data to datasets, also referred to as data drift management. 

As machine learning becomes prominent in an increasing amount of industries, its use must scale accordingly. Pioneering in the field by automating data pipelines, Lightly makes data management easy by integrating other data preparation tools and offering a centralised platform. This supports the trend towards data-centric AI, a concept elaborated by Andrew Ng. In Andrew’s words: “Data is food for AI”.

Today, Lightly is used by Fortune 500 companies and startups in autonomous driving, robotics, and video analytics. It plans to increase headcount and expand to the US.