Nextdata is building data meshes for the enterprise | TechCrunch

3 min read


The concept of a data mesh has been around for a few years. At a high level, it’s a data platform architecture that allows users to access data without transferring it to one of two places: a data lake, or a centralized repository for storing data at scale; or a data warehouse, an enterprise system used for analyzing data from multiple sources.

A number of data mesh products exist, including from big vendors like Databricks and Snowflake. But they’re not all created equal. At least, that’s the assertion of Zhamak Dehghani, the founder of Nextdata, a startup creating a “data-mesh-native” platform to build and share what Dehghani describes as “data products.”

For one, some data mesh vendors are getting VC attention — including (but not limited to) Nextdata. Nextdata today announced that it raised $12 million in a seed investment led by Greycroft and Acrew Capital, which Dehghani says will be put toward developing its tooling and expanding hiring across Nextdata’s product, engineering and go-to-market teams.

Dehghani says that she launched Nextdata to solve the challenges around data sharing as they relate to AI and machine learning.

“Today, IT leaders have to either choose between speed of data-driven innovation at the cost of safety and security; or slow data-driven innovation with security,” Dehghani told TechCrunch in an email interview. “The hype around AI isn’t rooted in reality for most enterprises today, as their data management technology still doesn’t enable the sharing of trusted data required.”

Prior to starting Nextdata, Dehghani was director of emerging technology at the tech consulting firm ThoughtWorks, where she evangelized the concept of the data mesh to her enterprise clients.

Inspired by her work at ThoughtWorks, Dehghani built a data mesh system — Nextdata OS — that allows users to create, share, discover and apply data products for analytics purposes. With ThoughtWorks, users can store data and metadata in a single container — the data product container — that other users can then manage or build into apps, websites and services.

Every Nextdata data product container has data governance policies “embedded as code.” These controls are applied from build to run time, Dehghani says, and at every point at which the data product is stored, accessed or read.

“Nextdata does for data what containers and web APIs do for software,” she added. “The platform provides APIs to give organizations an open standard to access data products across technologies and trust boundaries to run analytical and machine-learning workloads ‘distributedly.’ Instead of requiring data consumers to copy data for reprocessing, Nextdata APIs bring processing to data, cutting down on busy work and reducing data bloat.”

Nextdata claims to have pilots with Fortune 100 companies as well as “thousands” of users in its early access program. The near-term focus is expanding the startup’s team, which currently stands at around 10 people (four contractors and six full-time employees).

“With the immense hype around AI, there has never been more urgency for organizations to rethink their data systems to simplify the process of sharing, connecting and discovering enterprise data for analytics and AI in a decentralized and scalable fashion,” Dehghani said. “We’re building a team of visionaries and problem-solvers to lead the movement to responsibly unlock data for all.”


Source link