Databricks Data + AI Summit 2026: Enterprise AI Needs Customer Interaction Data

For four days in June, I was immersed in conversations about how large enterprises are preparing their data environments for a new generation of AI applications. More than 30,000 data and AI professionals gathered at the Moscone Center in San Francisco for Databricks Data + AI Summit 2026, where discussions focused on the platforms, governance practices, and operating models that help organizations put data to work across the business. 

The sessions that stayed with me most focused on the practical work behind AI inside large organizations. The technology has to reach the right business information, work within governance requirements, and support real work across many functions.

Given my role at Avaya, I kept coming back to where customer interaction data fits into that picture. Customer interaction data reveals how enterprise decisions are actually experienced by customers. It shows whether a policy change, operational update, or service decision achieved the intended outcome, making it an essential part of the enterprise data foundation that informs AI, workflows, and what happens next. 

Production AI Has to Work Across the Business 

One session that brought this into focus was CAVA’s presentation on Ask Astro, its production multi-agent supervisor on Databricks. The application lets employees ask business questions in plain language, such as which restaurants performed best or how loyalty redemption varies by region. Ask Astro routes each question to the appropriate source across finance, operations, customer analytics, HR, Jira, and web search before synthesizing an answer.

CAVA’s team focused as much on production readiness as model capability. Ask Astro routes across four Databricks Genie spaces, Atlassian APIs, and web search. The application includes guardrails around sensitive topics and personally identifiable information, and the team discussed what it took to deploy through Databricks Apps and Slack, rebuild after early setbacks, and measure adoption, time savings, and return on investment. 

What stood out to me was the way CAVA brought specialized data domains into one employee experience. Customer interaction data adds something those other domains cannot: direct evidence of how enterprise decisions are experienced by customers. It provides the customer's view of the problem, the journey so far, and the work that may still need to happen. 

Enterprise AI Depends on Governance That Can Scale

Connecting information across business domains only works when an organization can govern that information at scale. A separate session on Siam Commercial Bank (SCB) brought that challenge into perspective. 

Founded in 1907, SCB is Thailand’s first bank. After restructuring under SCBX in 2021, the group grew to 13 core subsidiaries. I kept thinking about what it takes to make more than a century of banking history and data useful across an organization of that size while maintaining the governance, security, and quality each part of the business requires. 

The session described SCB’s use of Databricks and Unity Catalog to support enterprise-wide data governance and data quality. It also demonstrated “Chat with your Metadata,” a Genie-based capability that helps users discover and understand the data available across the platform. 

The example made the foundation beneath enterprise AI more concrete for me. 

People and AI both need to find the right information, understand what it represents, and use it within the right access controls. Metadata, data quality, and governance make that possible across a large organization.  

Customer Interaction Data Shows Where the Experience is Breaking Down

The SCB session also made me think about the customer interaction data that belongs in that same enterprise view. It shows what happens after a business decision reaches the customer.

Consider a hypothetical scenario. A company updates a billing policy, and its billing system shows that 98% of affected accounts were updated correctly. From the company’s perspective, the change is complete. The customer journey tells a different story.

• Billing sees the financial impact. Refund requests and payment disputes rise by 25% after the change takes effect.
• Digital sees customers looking for answers. Visits to the payment FAQ increase by 50%, questions about unfamiliar charges rise in the virtual assistant, and exits from the payment page increase by 20%. Some customers may have already searched Google or an external AI tool before reaching a company-owned channel.
• Service sees the work left unresolved. Calls and chats about the charge rose by 40%. One in five customers contacts the company again after the first conversation, and one in four requires a transfer or escalation.

Viewed together, those data points show how the business is landing with customers in the moment. The billing change worked as intended, yet customers still encountered an unfamiliar charge, sought answers, and made repeated attempts to get help.

This is why customer interaction data matters beyond the contact center. It gives billing, digital, and service teams a shared view of whether customers understood the change, found an answer, and got the issue resolved.

Bring Enterprise Data Back to the Customer Journey

The CAVA and SCB sessions made the connection to customer experience clear. A production AI application needs connected data, and a large organization needs governance that makes that data usable. The customer journey is another place where the same model matters. 

Avaya Infinity is Avaya’s customer experience platform for connecting customer interactions, enterprise data, AI, and workflows across the journey. With Delta Sharing, Infinity can make live customer interaction data available in the Databricks lakehouse without creating extra copies. Enterprise data activation can then bring relevant account, policy, service, and operational information back into the interaction. 

In the billing example, an agent or AI assistant could work with account details, policy information, recent self-service activity, and interaction history. Real-time orchestration can use that context to guide routing, agent assistance, workflows, or AI-assisted action.

Connected data matters when it helps the business act before the customer has to ask again. 

Customer Experience Belongs in Enterprise Data Strategy 

I left Databricks Data + AI Summit with a clearer view of what customer experience teams need from an enterprise data strategy. Customer interaction data shows whether a business decision worked for the people it was meant to serve. Paired with billing, service, product, and operational data, it helps the teams responsible see what customers experienced and what needs to change. 

Avaya Infinity helps enterprises connect that context to data, AI, and workflows across the customer journey so teams can understand what is happening and move the work forward while it still matters.

Explore Avaya Infinity