Databricks and Infosys: AI Held Back by Silos of Data

AI Held Back by Dispersed Data

Artificial intelligence (AI) is at the heart of modern business strategies, but a persistent obstacle hinders its widespread adoption: the state of data. While AI tools are appealing due to their simplicity, their deployment requires a unified and governed data infrastructure, a challenge that many companies struggle to overcome. Bavesh Patel, Senior Vice President at Databricks, emphasizes that the effectiveness of AI heavily relies on the quality of internal data. In many organizations, this data remains scattered across legacy systems and siloed applications, making it difficult to produce reliable outcomes. “In reality, the major differentiating factor for most organizations is their own data and the third-party data they can add to it,” says Patel.

The Need for a Unified Data Architecture

For AI to deliver real added value, it is crucial to consolidate data into open and accessible formats. Patel warns that without this foundation, companies risk developing “terrible AI.” A unified data architecture would allow for the combination of structured and unstructured data while maintaining real-time context. This requires moving beyond siloed SaaS platforms and disconnected dashboards. Companies must turn to an open data architecture capable of preserving real-time context and enforcing strict access controls. When the foundations are laid correctly, organizations can move towards measurable outcomes, unlocking efficiency gains, automating complex workflows, and even launching new lines of business.

The Importance of AI Literacy

Rajan Padmanabhan, Chief Technology Officer at Infosys, highlights the importance of linking AI to business metrics. Successful companies integrate AI into their strategic decisions, using governance frameworks to assess the effectiveness of initiatives. “We see this great opportunity with AI literacy among business users, where they are very eager to understand how they should think about AI,” adds Patel. This involves knowing the elements and building blocks necessary from a technological, training, and empowerment perspective. Rather than treating AI initiatives as isolated innovation projects, leading companies directly link AI deployment to business metrics, using governance frameworks to determine what generates results and what should be quickly abandoned.

Towards a New Era of AI

The future of enterprise AI hinges on transforming fragmented data into a strategic asset. Patel and Padmanabhan envision an evolution of AI agents into autonomous operators capable of managing complex workflows. Companies that lay the right foundations now will be the ones to fully leverage these advancements. “What we see as a new way of thinking is moving from an execution or engagement system to an action system,” notes Padmanabhan. This is the new path we see ahead of us. The future of AI in business will be determined by companies' ability to transform fragmented information into a strategic asset capable of driving both smarter decisions and entirely new ways of operating.