From movie recommendations to chatbots, AI-powered applications are deeply embedded in the modern web experience. But building the same features for a so-called Web3 business—say, adding an Amazon-like product recommendation feature to an NFT marketplace—is complicated and expensive.
Seattle-based Spice AI, which announced Wednesday it raised a $13.5 million seed round, aims to fix that.
“Think about a bunch of the services you would use today — Netflix, YouTube, Amazon — things that your grandmother probably used. Their experiences are really driven by recommendations,” Spice co-founder and CEO Luke Kim told GeekWire. “But it’s very difficult to build those same experiences in the Web3 space because the infrastructure for that doesn’t exist.”
Spice AI plans to deliver that infrastructure through its own Web3 data warehouses. The startup indexes all the data from the major blockchain networks – such as Bitcoin, Ethereum and Solana – and then stores and distributes it on a platform. Customers can then access this data to build AI-powered applications for their own blockchain companies.
“Spice AI builds the foundational platform for Web3, so developers don’t have to invest in building their own infrastructure and in data engineering that is complex and expensive, and frankly hard to achieve,” Madrona partner Aseem Datar said in a statement .
Use cases vary across a wide variety of verticals. Web3 companies in industries ranging from financial services, fraud detection and retail use the service to build out their own AI-powered applications. The platform, which is currently in the beta phase, has around 10 customers.
Kim founded the company in 2021 together with Phillip LeBlanc. The duo worked together at Microsoft on Azure technologies including Dapr and Azure Active Directory. Kim also helped create the incubation team at Azure. Spice AI currently has fewer than five employees, but it will use some of its fresh cash to grow to a workforce of 10 engineers by the end of the year.
The startup’s relatively large seed round comes at a time when many venture capitalists are pulling back on their bets on crypto-related startups. VCs invested just $6.76 billion in crypto companies in the period ending June 30, representing a decline of more than 31% from the previous quarter, according to PitchBook. Spice has raised $14.5 million in funding since 2021, including a $1 million pre-seed round in October.
Madrona led both funding rounds; other investors include Australia-based Blackbird Ventures, Basis Set Ventures and Founders’ Co-op, among others. The cap table also features a trio of notable tech leaders, such as former GitHub CEO Nat Friedman, current GitHub CEO Thomas Dohmke, and Microsoft Azure CTO Mark Russinovich.
According to Kim, the startup’s biggest competitors are experienced developers at Web3 companies who can retrieve the data themselves, store it using other software platforms such as DataGrip or Snowflake. There are also a bunch of startups looking to collect and store Web3 data, including Dune Analytics, which has a query system, and Nansen AI, a blockchain analytics platform that focuses mainly on financial transactions.
Blockchain networks have huge data sets, which means that Spice AI will take on huge overhead costs to store the information. However, these costs will eventually be amortized to the customers.
“We can get pretty good economies of scale out of it,” Kim said.
One of the most important hurdles the startup will face is keeping up with growing pains in the nascent crypto space. Solana, for example, has dealt with full or partial outages at least seven separate times in the past 12 months, Cointelegraph reported. Ethereum, meanwhile, is merging its platform into a new authentication process in the coming week.
“You have to do a bunch of work and test around, and make sure all that technology works and continues to work,” Kim said. “The space is changing very quickly.”
