Visual Layer provides tools to help data scientists and ML engineers detect duplicates, broken files and mislabeled images in very large visual datasets. At its core is the open-source fastdup project; the company’s system can analyze hundreds of millions of images without relying on expensive GPUs, clustering and visualizing potential issues. Visual Layer offers fastdup as a service with additional enterprise features layered on top. The company was co-founded by Danny Bickson (CEO) and Amir Alush (CTO), with Carlos Guestrin among the co-founders, and the team has experience from Apple, Turi and Brodmann17. Early users include Meesho (which used fastdup to improve a 200 million-product image gallery), John Deere, Honeywell, Winnow and Nuvilab. Visual Layer’s research has highlighted large-scale dataset problems, for example finding over a million duplicate pairs in the ImageNet-21K set.

Total raised

$7M

Funding rounds

1

Latest round

Seed

Latest activity

May 2023