DataHub gets $35M in funding to provide the context needed for AI reliability and safety

Open-source metadata startup DataHub, owned and operated by Acryl Data Inc., said today it has closed on a $35 million Series B funding round led solely by Bessemer Venture Partners.

The round brings the company’s total amount raised to more than $65 million, and will enable it to accelerate the development of its flagship context management platform, which provides teams with the ability to discover, observe and control data for artificial intelligence models and AI agents.

DataHub says in its pitch that enterprises currently face major headaches when it comes to accessing, securing and maintaining the reliability of their data and AI supply chains. As such, their AI initiatives struggle with missing context that prevents both humans and AI machines from working effectively with enterprise information.

The challenges are threefold, DataHub says. Human workers can’t easily find the relevant datasets they’re looking for, while data engineers lack the visibility they need to prevent disruption when making changes, and governance teams struggle to keep track of who has access to sensitive data sources.

These problems also extend to AI models, which are required to understand when new information comes available to refresh their predictive capabilities and which data is trustworthy. In addition, they also struggle to analyze schema changes automatically.

DataHub’s platform is designed to address all of these challenges by providing the underlying real-time metadata that’s required to interact with data assets with full context awareness.

It’s the creator of a modern data catalog that’s designed to simplify metadata management, discovery and governance, enabling any users to explore and understand their company’s data, track lineage, profile datasets and establish data contracts with ease. The platform aims to help developers tame the complexity of their constantly evolving data ecosystems, while enabling workers to leverage the value hidden within their organization’s data assets.

The startup has created a novel, event-driven metadata architecture that provides real-time visibility into changing data estates. It says it’s a highly scalable and extensible platform that boasts flexible deployment options, ranging from single-node to cloud-hosted, hybrid and decentralized deployments.

DataHub believes that the rapid adoption of AI in the enterprise has increased the importance of having comprehensive visibility into the data ecosystem, as it’s the key to ensuring greater trust and reliability. As such, there’s a need to move beyond traditional data cataloging, to a machine-scale world where AI becomes the power users of data.

“With the shift toward business-critical AI and customer-facing predictive applications, enterprises need robust metadata management to ensure AI systems can reliably work with data,” said co-founder and Chief Technology Officer Shrishanka Das. “DataHub provides the context that AI systems need to understand data lineage, quality, and semantics – enabling organizations to unlock the full potential of their AI investments.”

The startup said the biggest advantage of its platform is its extensible nature, which allows it to unify capabilities across data discovery and observability and support the unique needs of AI governance. It claims to have had some significant customer wins in recent years too, with the likes of Apple Inc., Foursquare Inc., Netflix Inc., Salesforce Inc.’s Slack and Pinterest Inc. all using its platform, helping to drive revenue growth of more than sixfold over the last two years.

Looking forward, DataHub is planning to invest significantly in its open-source community, which already boasts more than 13,000 members, and accelerate its research and development, focusing chiefly on its AI governance and context management capabilities. The company also plans to scale its go-to-market and customer success teams.

Bessemer Venture Partners’ Lauri Moore said metadata has emerged as the missing link that will enable organizations to evolve from human-scale to machine-scale data analytics.

“DataHub is uniquely positioned to address this critical need with its schema-first, event-oriented architecture, which brings model and data context and control into a single pane of glass,” she said. “Enterprises will use DataHub to develop AI safely, in a way that respects user privacy and ensures that people, models and AI agents only access the data and context when and where they are supposed to.”

Image: SiliconANGLE/Dreamina