NEW YORK –Grid, a company pushing new state-of-the-art standards in artificial intelligence (AI), announced that it has raised $18.6 million in Series A funding led by Index Ventures with participation from Bain Capital Ventures and Firstminute. Grid enables companies of all sizes to train state-of-the-art AI models on hundreds of cloud GPUs and TPUs.
Grid’s breakthrough technology allows machine learning engineers, data scientists and AI researchers to focus on solving problems and building products at the scale available to only the most sophisticated AI companies in the world.
“Grid is here to teach, train and supercharge machine learning engineers, data scientists and AI researchers, so they can work with maximum efficiency on exactly the way they want, instead of spending all their time managing hardware or solving engineering problems,” said Grid cofounder William Falcon. “Our main goal is for the product to be user friendly so companies don’t have to worry about the machines they are running or clusters, and instead can focus on delivering value for their customers.”
While doing his PhD at NYU and part-time at Facebook AI Research, Falcon created PyTorch Lightning, one of the fastest growing machine learning frameworks in the world, amassing over 600,000 downloads in the last 12 months. Lightning gives AI researchers, data scientists and machine learning engineers the flexibility to iterate on research ideas faster without needing to be expert research engineers. Lightning is used by 400+ organizations, ranging from small startups to Fortune 500 companies and many top academic and AI research labs. Falcon and Grid cofounder Luis Capelo, who most recently served as the head of Machine Learning at Glossier, saw an opportunity to use PyTorch Lightning to reduce the distance between deep learning research and its practice in real-life businesses. With this $18.6 million investment, they are building Grid’s core technologies and growing Grid’s New York-based team.
PyTorch Lightning’s core innovation is around decoupling all the code that is required to define a full deep learning model from all the code that is required to run on hardware—in short, pivoting the focus away from the engineering and into solving the science and business problems. Grid’s approach has been so successful that other projects such as Keras, Tensorflow and Huggingface have adopted similar paradigms in their own frameworks. As a platform for training models, Grid extends the power of PyTorch Lightning and makes it possible for any company to have access to state-of-the-art AI services on par with those of Google and Facebook.
“The unique combination of the explosive popularity of PyTorch Lightning and the deep technical expertise of Grid’s founding team will enable Grid to have a profound impact on the ability of all companies, regardless of resources, to take advantage of cutting edge machine learning technology,” said Sarah Cannon of Index Ventures.