0%
Stay Updated
PRODUCTS_
RESOURCES_
© 2025 Provably. All rights reserved.
Provably V1.3 is live! While still in Alpha, this release marks the debut of the Verifiable Analytics Network, connecting Data Owners with Insight Seekers. Privacy preserving collaboration remains central to Provably, so now Data Owners can create data collections accessible to specific parties or made public. Insight Seekers can run verifiable analytics across a broader range of private data repos (it is not 1 to 1 only anymore). Additionally, Provably’s set of SQL aggregate queries has expanded (see details below).
Since our initial V1 launch in August, we’ve spoken with many builders to understand the challenges they face. While numerous tools exist for data science within development teams and organizations, there is a new, emerging actor in need of private data—AI agents. In upcoming releases, we’ll release more tools for these new species 🙂 (check out our roadmap). For now, we’re excited to share the foundational work we’re laying for the AI era.
Provably continues to demonstrate the potential of Zero Knowledge for Big Data. Previously, data owners could connect Provably to their databases and run private, verifiable queries using SUM and COUNT SQL aggregations. This release adds STANDARD DEVIATION, VARIANCE, and MEDIAN, expanding the analytical capabilities of Provably.
As one of the few teams pioneering ZK for Big Data analytics, we’re gaining practical experience in supporting larger datasets and improving infrastructure and user experience. Stay tuned—exciting updates on infrastructure and core protocol developments are on the way!
For user experience, we’ve introduced Data Collections—cards that showcase curated datasets from data owners with detailed descriptions and instant query capabilities. Inspired by catalogs on Kaggle, Databricks, and Snowflake, we aimed to streamline access to data insights. Now, if you find an interesting collection, you can query it instantly, getting insights without delay. We’ve also built a set of live basketball collections with results updated within the last 24 hours, so join the Alpha, run some queries, and see what you find! APIs are also opening up to allow other interfaces to run analytics on these collections.
Data collections can be shared publicly with all Provably users or in Limited Access mode, where Data Owners invite specific users by email to run analytics on their data. This feature marks a step towards multi-layered data and AI ecosystems, allowing selective sharing with trusted networks or collaborators. It’s a first step towards creating what we call Verifiable Analytics Networks.
Provably invites you to interact with our Alpha program and explore the future of collaborative, privacy-guaranteed computing. Join us soon on Discord, where we’ll be welcoming your insights and questions.
We talked to many builders, data owners, AI agent designers and last but not least Data Acquisition Scouts and we heard this a few times. What differentiates Provably from some of the other Web3 projects that are announcing verifiable SQL 🙂? While we do not want to go into the technical details here (there will be time for that in a more technical blog), we just want to leave a message:
YOUR DATA
YOUR DATABASE
YOUR INFRASTRUCTURE
Provably does NOT require you to replace your secure storage infrastructure. Plug and go!
And here’s a sonnet (generated by an AI agent):
In your domain rests all your precious data, Secure and sovereign, never to migrate. Your infrastructure stands, unswayed, no traitor, Provably steadfast, it guards without debate.
No urge to copy, clone, nor replicate, Your storage is safe, no need to re-create. A fortress firm, where trust is not a wager, And privacy upheld without a waiver.
For your data's realm, no siege lays claim, No covert breach, no silent, stealthy game. Your digital expanse remains the same, Secure in its foundation, free from blame.
Thus hold your lines, your protocols adhere, Provably intact, your data's refuge clear.
