In brief
- Space and Time has launched Space and Time Enterprise, a platform that lets businesses prove calculations on private data are correct without exposing the underlying records.
- Launch partners include Maple Finance, Indomobil, Trenez and The Data Foundation, and businesses can run the software on their own servers without operating a blockchain.
- AI agents "don't need unrestricted access to the records behind every calculation," Nate Holiday, CEO of Space and Time developer MakeInfinite Labs, told Decrypt.
Space and Time, a Microsoft-backed blockchain data project, has launched a platform designed to let businesses use private data in AI tools without handing over the underlying records, the company announced on Monday.
Space and Time Enterprise lets companies prove that calculations made on their private data are correct, so that AI applications, auditors and other stakeholders can check the results without accessing the database itself. Launch partners include onchain asset manager Maple Finance, Indonesian vehicle maker and distributor Indomobil, Oracle software consultancy Trenez and AI data network The Data Foundation.
"Businesses want AI agents to execute workflows using that information, but those agents don't need unrestricted access to the records behind every calculation," Nate Holiday, CEO of MakeInfinite Labs, the original developer of Space and Time, told Decrypt.
The stakes are high, he argued. "The potential cost to a business is the overall value of the business," Holiday said. "Giving AI tools unrestricted access to data creates another avenue of exposure."
Holiday said the company is working with customers in industries including construction, engineering and manufacturing that want to rebuild their systems around AI, so that agents, reporting tools and compliance teams can work from the same business records and check the calculations they rely on.
Proof of SQL and the SXT token
The platform is built on Proof of SQL, Space and Time's patented technology for proving that database queries were computed correctly. It already works with Google Cloud's BigQuery, and businesses can now run it on their own servers without operating a blockchain, buying specialized hardware or relying on an outside proving service, according to the company.
With GPU acceleration, supported queries over millions of rows can be run and proven in under a second, Holiday said, though actual performance depends on the query, hardware and dataset.
The technology proves that a calculation was carried out correctly on a given set of records, but it cannot tell whether those records were accurate to begin with. "An incorrect entry still needs to be caught through the business's data controls," Holiday said, "but verification gives those controls a much stronger foundation: evidence of which data was used and whether it was processed correctly."
Space and Time raised $20 million in 2022 in a strategic funding round led by M12, Microsoft's venture fund. Last year, Grayscale launched a trust giving investors exposure to SXT, the token that powers Space and Time's blockchain. Earlier this year, the company launched Dreamspace, an AI-powered tool built with Microsoft's Azure AI services that lets users create and deploy onchain apps without writing code.
Space and Time also runs SXT Chain, a decentralized network that supplies verifiable data to DeFi, tokenized assets and stablecoins. Asked whether enterprise customers would create demand for SXT, Holiday said businesses could deploy the product in their own environment, and that the network would come into play as their applications "need to connect to onchain markets," with those workloads paid for through Space and Time.
The company's cryptography is also used elsewhere. Microsoft's Nova project includes Blitzar, Space and Time's cryptographic acceleration library, and the team's work on a commitment scheme called Dory contributed to a16z crypto's Jolt proof system, Holiday said.
Daily Debrief Newsletter
Start every day with the top news stories right now, plus original features, a podcast, videos and more.