According to a new report from Databricks, the landscape of enterprise AI is undergoing a seismic shift. Organizations are moving beyond simple chatbots and pilot programs, embracing intelligent, agentic systems that redefine operational workflows.

Generative AI’s initial wave was marked by high expectations but often limited utility. Technology leaders grappled with isolated tools that failed to deliver transformative business value. However, fresh telemetry data from Databricks indicates a turning point. The market is maturing, and the focus is now on autonomous agents capable of planning and executing complex tasks.

Data drawn from over 20,000 organizations—including 60 percent of the Fortune 500—reveals a rapid migration toward architectures where AI models don't just retrieve information but independently orchestrate workflows. This evolution represents a fundamental reallocation of engineering resources, with multi-agent workflow usage on the Databricks platform surging by 327 percent between June and October 2025.

The Rise of the 'Supervisor Agent'

Central to this growth is the concept of the ‘Supervisor Agent’. Instead of relying on a monolithic model to handle every request, a supervisor acts as an orchestrator. It breaks down complex user queries and delegates specific tasks to specialized sub-agents or tools, much like a project manager in a human team.

Since its launch in July 2025, the Supervisor Agent has quickly become the dominant agent use case, accounting for 37 percent of all usage by October. This structure mirrors effective organizational hierarchies: a manager ensures execution without performing every task personally. Similarly, a supervisor agent manages intent detection, compliance checks, and routing to domain-specific tools.

While technology companies are currently leading this charge—building nearly four times more multi-agent systems than any other industry—the utility of agentic AI extends far beyond tech. Financial services firms, for example, are deploying multi-agent systems to handle complex document retrieval and regulatory compliance simultaneously, delivering verified client responses without human intervention.

Infrastructure Under Pressure: The Real-Time Reality

As AI agents graduate from answering questions to executing tasks, the underlying data infrastructure faces unprecedented demands. Traditional Online Transaction Processing (OLTP) databases, designed for human-speed interactions, are being pushed to their limits. Agentic workflows invert these assumptions, generating continuous, high-frequency read and write patterns.

The scale of this automation is staggering. Two years ago, AI agents created a mere 0.1 percent of databases. Today, that figure sits at an impressive 80 percent. Furthermore, 97 percent of database testing and development environments are now spun up by AI agents, allowing developers to create ephemeral environments in seconds rather than hours.

Contrary to the big data legacy of batch processing, agentic AI operates primarily in the "now." The report highlights that 96 percent of all inference requests are processed in real-time. This shift is particularly evident in sectors where latency correlates directly with value, such as healthcare and finance.

The Multi-Model Standard and Vendor Independence

Vendor lock-in remains a persistent risk for enterprise leaders. To mitigate this, organizations are actively adopting multi-model strategies. As of October 2025, 78 percent of companies utilized two or more Large Language Model (LLM) families, such as ChatGPT, Claude, Llama, and Gemini.

The sophistication of this approach is increasing. The proportion of companies using three or more model families rose significantly from 36 percent to 59 percent in just two months. This diversity allows engineering teams to route simpler tasks to smaller, cost-effective models while reserving frontier models for complex reasoning tasks.

Governance as an Accelerator

Perhaps the most counter-intuitive finding for many executives is the relationship between governance and velocity. Often viewed as a bottleneck, rigorous governance and evaluation frameworks function as accelerators for production deployment.

Organizations using AI governance tools put over 12 times more AI projects into production compared to those that do not. Similarly, companies employing evaluation tools to systematically test model quality achieve nearly six times more production deployments. Governance provides the necessary guardrails—such as defining data usage and setting rate limits—giving stakeholders the confidence to approve deployment.

The Value of 'Boring' Automation

While autonomous agents often conjure images of futuristic capabilities, the current enterprise value lies in automating routine, mundane tasks. The top AI use cases vary by sector but focus on solving specific, practical business problems:

• Manufacturing & Automotive: 35% of use cases focus on predictive maintenance.
• Health & Life Sciences: 23% of use cases involve medical literature synthesis.
• Retail & Consumer Goods: 14% of use cases are dedicated to market intelligence.

Furthermore, 40 percent of the top AI use cases address practical customer concerns such as support, advocacy, and onboarding. These applications drive measurable efficiency and build the organizational muscle required for more advanced agentic workflows.

For the C-suite, the path forward involves less focus on the “magic” of AI and more on the engineering rigour surrounding it. Competitive advantage is shifting back towards how companies build, rather than simply what they buy. Open, interoperable platforms allow organizations to apply AI to their own enterprise data, creating long-term differentiation in highly regulated markets.