Products & Tools

Commvault and Databricks turn backed-up data into governed AI fuel

Commvault Data Activate for Databricks moves protected enterprise data into Databricks as governed tables, after research found 70% of AI leaders can use under half their unstructured data.

Commvault & Databricks launch governed data offering
Commvault & Databricks launch governed data offeringcookieone / pixabay
By Sophie Lindqvist6 min read

Why it matters

  • Commvault announced Data Activate for Databricks on 17 November 2026.
  • Commvault-cited research: 70% of data and AI leaders say less than half of their unstructured data can be found and used for AI.
  • The offering shares data through OpenSharing as governed tables managed via Unity Catalog.
  • Early access is planned in select Azure regions.
  • Commvault is already a certified Databricks partner providing data protection for the Databricks Data + AI Platform.

Commvault has announced Commvault Data Activate for Databricks, a joint offering that moves protected enterprise data into the Databricks platform as governed tables ready for analytics, AI applications and AI agents. The announcement, made on 17 November 2026, extends an existing partnership between the two companies beyond backup and recovery into data preparation and governed sharing.

The stakes here are straightforward. Commvault cited research showing that 70% of data and AI leaders say less than half of their unstructured data can be found and used for AI. That figure explains why two established vendors are building a product around data that companies already own but largely cannot reach.

What does the product actually do?

The mechanics matter more than the branding. The offering is built on Databricks and prepares data managed by Commvault before sharing it through OpenSharing as governed tables. Those tables can then be managed through Unity Catalog, Databricks' governance layer, and used across analytics and AI workloads inside the platform.

The point of that pipeline is to avoid a persistent integration headache. Instead of rebuilding connections to each underlying source system — a file share here, an application repository there — organisations can select data already protected by Commvault and make it available in Databricks without establishing a separate link to every original source.

The structure is also designed to carry business context with the data as it moves. According to Commvault, that context includes source details, ownership, access permissions and lineage, while also preserving retention and recoverability. In plain terms: the data arrives in Databricks knowing where it came from, who owns it, who may see it, and how long it must be kept.

Takeaway: the product tries to move enterprise data into AI workflows without stripping it of the governance metadata that makes it trustworthy.

Why is unstructured data the focus?

Much of the offering's focus sits on unstructured data — documents, files and content spread across multiple systems rather than neatly ordered databases. Many companies store this material across fragmented environments, and it rarely travels well into analytics platforms.

The 70% figure from Commvault's research describes the scale of the problem. If most data and AI leaders admit that under half of their unstructured data is even findable and usable for AI, then a large share of corporate information is effectively invisible to the systems companies are building. That is a direct constraint on enterprise AI projects, which depend on access to trusted internal data that remains subject to governance rules.

Many businesses hold large volumes of information in file systems, applications and repositories that are protected for resilience purposes but not readily available for analysis or AI models. Backup estates, in other words, hold the data — they just were never designed to serve it up for modelling.

Takeaway: the bottleneck for enterprise AI is often not compute or models, but governed access to data that already exists inside the company.

How do the companies frame it?

The two vendors frame the partnership from complementary angles, and their statements carry the voice of each side's commercial interest.

Alan Atkinson, Chief Business Development Officer at Commvault, positioned the offering as an evolution of what Commvault already does. "Commvault protects the data enterprises depend on," Atkinson said. "Together with Databricks, we aim to make more of that trusted data available for AI while preserving the context that gives it value. This is about turning protected data into an active business asset."

That last sentence is the thesis of the product. Backup data, historically a passive insurance policy, becomes an input to analytics and AI.

Databricks, meanwhile, framed the deal in terms of making enterprise data more useful inside its own environment. "Customers tell us their data is most valuable in Databricks when that data's context travels with it," said Stephen Orban, SVP of Product Partnerships and Ecosystem at Databricks. "Working with Commvault can expand the governed, trusted data available across Databricks, giving search, analytics and AI applications more to work with across the business."

Orban's comment signals what Databricks gains: more governed data flowing into its platform, feeding not just dashboards but search and agentic AI workloads as well.

Takeaway: Commvault wants its protected data to be an active asset; Databricks wants more trusted data inside its platform — the product serves both ambitions at once.

What are the three core functions?

The companies position the product around three main functions:

  • Bringing additional governed enterprise data into Databricks from information already protected by Commvault, without creating separate source-by-source links.
  • Preserving business context alongside the data, including source details, ownership, permissions and lineage.
  • Keeping information current over time by reflecting new, changed and deleted content, so that analytics outputs and AI systems remain aligned with source data.

The third function deserves attention. Stale data is a quiet failure mode for AI systems — models trained or grounded on outdated content produce confidently wrong results. By syncing new, changed and deleted content, the offering tries to keep AI outputs aligned with the actual state of enterprise information.

Takeaway: the pitch is not just "move the data" but keep it governed, contextual and current as it moves.

What changes in the Commvault–Databricks relationship?

The relationship between the two companies is not new. Commvault already provides data protection for the Databricks Data + AI Platform and is a certified Databricks partner. What this announcement changes is the scope: the partnership now covers data preparation and governed sharing for AI and analytics, not just backup and recovery.

That shift reflects a wider push by technology suppliers to address one of the main constraints on enterprise AI: getting trusted internal data out of protected silos and into active use. Rather than treating backup estates solely as a resilience function, vendors are increasingly presenting them as controlled sources of enterprise information that can be reused for operational and analytical work.

The emphasis on context and governance also reads as an answer to customer concerns about the provenance and control of data used in AI systems. Companies building AI agents and applications face growing pressure to know exactly which data those systems touch, who owns it, and under what rules it may be used.

Takeaway: backup vendors are recasting themselves as data suppliers for AI, and this deal is a concrete example of that repositioning.

When and where will it be available?

Early access to Commvault Data Activate for Databricks is planned in select Azure regions. That points to an initial cloud-specific launch rather than a broad release across all environments, so organisations running multi-cloud or on-premises estates should not expect immediate availability.

What should customers watch?

For customers already using both platforms, the main question is practical: does the new service genuinely reduce the integration effort needed to expose protected data in a governed form? The stated goal is that organisations can select data protected by Commvault and make it available in Databricks for analytics, AI applications and agents without a separate connection to each original source.

The answer will decide whether this becomes routine infrastructure or another layer of integration work. Early access in select Azure regions will give the first evidence, and a broader rollout — if it follows — will show whether turning protected data into an active business asset works as smoothly as the announcement suggests.

Source: itbrief.com.au

Share this article:

More from Sophie Lindqvist

Sophie Lindqvist

Show full bio

Staff writer covering regulation, policy and their impact on artificial intelligence and machine learning at AI In Context.

749 articles

Related articles

  1. Commvault and NVIDIA to Make Protected Enterprise Data Usable for AI
  2. Commvault and Zscaler Team Up to Govern and Recover Enterprise AI Agents
  3. Databricks Maps Three Enterprise Paths for OpenAI Tools on Governed Data
  4. Cohesity Expands Backup, Recovery and AI Data Access Across Microsoft Stack
  5. Databricks’ Genie App for Slack Is Now Generally Available

« Previous articleNext article »