A new zero-knowledge proof (ZK-proof) technology, proof of SQL, has been introduced. It promises to slash the traditional 30-minute wait time for Structured Query Language (SQL) queries to sub-second speeds. 

The drastic reduction in time can lead to substantial efficiency gains in applications that require real-time data processing. 

Space and Time (SxT) — a verifiable compute layer for artificial intelligence (AI) and blockchain — revealed the new open-source ZK prover for processing data on GitHub on June 12. 

Related: StarkWare launches $1M research fund for ZK Bitcoin scaling

ZK-proof breakthrough

Speaking to Cointelegraph, Jay White, SxT co-founder and head of research, explained the benchmarks used to measure the sub-second performance:

“We ran benchmarks with various queries, such as filters, GROUP BYs, aggregations and other standard SQL operations against large data tables containing over 100,000 rows. The latest benchmarks show that proof of SQL can prove against 100,000 rows in under a second, one million rows in 1.2 seconds, and 100 million rows in approximately one minute.”

The proof of SQL was released in alpha to a select group of SxT customers in August 2023 and is now available to the public on GitHub. 

The publicly available cryptographic tool will allow community members to run trustless queries on SxT, and developers can download the repository directly from GitHub.

“Our team pioneered sub-second ZK-proofs so that smart contracts and AI agents can ask questions about a chain’s activity, as well as off-chain data, and receive back trustless SQL query results on-chain during a transaction without having to wait for 30-minute proof times.”

Related: Polygon Labs acquires Toposware, pushing total ZK investment to $1B

ZK-proofs go open-source 

The new ZK-proof is part of SxT’s broader strategy to improve the security and performance of decentralized applications (DApps) and encourage community collaboration.

White added: “Our team cares deeply about UX [user experience], which is part of why we chose to hyperfocus on SQL [...] to make integrations as seamless and easy as possible for developers.

“We’re currently working on integrating proof of SQL in the Chainlink network, [and] working with the zkSync ecosystem on a solution allowing developers to run SQL queries natively from their smart contract with proof of SQL.” 

The delay inefficiency hinders the scalability and performance of DApps relying on real-time data verification. 

According to a press release shared exclusively with Cointelegraph, proof of SQL “allows developers to connect analytics directly to smart contracts, opening up a wealth of powerful new use cases and business logic on blockchain technology.”

Related: Matter Labs drops ZK trademark applications after industry backlash

Tackling scalability challenges

White explained how proof of SQL will scale with increasing data volume, more complex queries and future enhancements to offset bottlenecks:

“Proof of SQL scales linearly. Cluster-scale proofs are on the near-term roadmap, meaning even faster proofs over even larger tables. Other upcoming enhancements include expanded SQL support, particularly multi-table queries (like JOIN) and subqueries and a novel commitment scheme.”

The integration of ZK-proofs in the crypto industry could have significant implications. White explained what kind of use cases he thought this development in ZK could have:

Our team firmly believes AI agents will transact primarily on-chain within the next few years. Accurate, untampered, provably correct data through proof of SQL will fuel this on-chain agent revolution. Every app in the world uses a database, and [...] we see a future where every database uses proof of SQL.

The push for ZK technology is advancing ever-faster, evidenced in the recent launch of StarkWare’s $1 million research fund to roll out ZK-scaling for Bitcoin.

The initiative aims to scale both Ethereum and Bitcoin simultaneously through ZK technology.

Magazine: Longevity expert: AI will help us become ‘biologically immortal’ from 2030