Graft Gets Rolling With $4.5M Pre-Seed Round

May 12, 2022

Graft emerged from stealth to put modern AI within reach of every organization. Graft distills the AI infrastructure and workflows that top companies are converging around into an all-in-one package that is easy to use, regardless of machine learning proficiency or existing infrastructure.

“Other vendors sell you a piece of the puzzle, but that’s only useful if you have the rest of the pieces and a team to assemble them. You would spend lots of time and money setting up the machinery of AI instead of investing in differentiating business value. We started wondering, what if someone provided the power of modern AI without the prerequisites? That’s how Graft was born,” said Adam Oliner, co-founder and CEO of Graft.

Modern AI companies like Google and Meta leverage their AI infrastructure to great effect for a breadth of use cases like search, recommendations, content moderation, personalization and customer experience. With that infrastructure, they can use many different types of data, including text, images, audio, video and graphs, and can go from idea to production success on the same day. At the same time, they devote hundreds of people and hundreds of millions of dollars to build and maintain it.

Organizations looking to acquire comparable AI capabilities face major obstacles. They need to hire an expensive team of ML engineers and data scientists, assemble production infrastructure using a dozen different open source libraries and cloud services, layer on tooling to ensure correct, compliant, efficient and reliable operation and then manage and maintain it all. It is an expensive, time-consuming process.

Co-founder and CEO Adam Oliner faced this first-hand when he was head of machine learning at Slack. There they wanted to do what modern AI companies were doing with technologies like pre-trained models and embeddings but there was no easy path. They had to build and integrate the key infrastructure mostly from scratch. Once it was in place, they were able to deploy substantially better search and recommendations throughout the product, while also unlocking a variety of other use cases.

Oliner went on to found Graft with Maria Kazandjieva (former director of engineering at Netflix), Eric Schkufza (former principal scientist at Amazon), Mick Hakobyan (distinguished tenures at Meta, Netflix and Microsoft), Irina Calciu (former senior researcher at VMware) and Brian Calvert (former machine learning platform technical lead at Cruise Automation) to make modern AI accessible to organizations at any stage of their AI journey.

Graft removes typical AI barriers to entry: expertise, infrastructure and cost. Instead of cobbling together makeshift solutions over months or years, organizations can focus on activities that yield true, differentiating value, like investing in data and label quality. Everyone in the organization from business analysts to product managers can apply the resulting intelligence using powerful SQL extensions and straightforward production APIs.

Graft Raises a $4.5 Million Pre-Seed Round

Today Graft also announced that it has raised a $4.5 million pre-seed round led by GV. NEA, Essence VC, Formulate Ventures and SV Angel also participated. Graft plans to use the money to continue growing the team, building the product and bringing it to market.

"Most enterprises have huge and growing data repositories and very few are capturing all the insights available in these troves. Graft's approach to modern AI will make this process infinitely easier for businesses that crave deeper insights without dedicated machine learning talent," said Erik Nordlander, General Partner at GV. "Adam Oliner is a uniquely brilliant technical leader with an ability to execute, and we're excited to welcome him and the Graft team to the GV portfolio."