Solving the Problems That Accompany API Sprawl With AI

API sprawl hides both security dangers and missed opportunities. If an organization has more APIs than it can easily keep track of, those APIs can easily become entry points for malicious actors.

And if those APIs are underutilized, they’re not making the most of their potential to generate revenue.

“APIs run the core business process you want to scale,” said Neeraj Nargund, who oversees the strategic partnerships for the automation brand at IBM, in this episode of The New Stack Makers.

”And when you scale it across the enterprise with so many different business functionalities, processes, etc., you want to have automated discovery. You want to have observability. You want to have governance.”

At scale in an enterprise environment, that governance is particularly vital. “What are the endpoints? Are there shadow API remediations?” Nargund asked. “That’s thousands of APIs that you build. You need a better governance at scale.”

Monetizing the data APIs generate and turning that data into products is another need that’s only grown in the AI era, he said. “How do you package the data and the insights, and how do you help curate price and segment usage?” he asked. “All these become capabilities that you need.”

How AI-Infused ‘Smart APIs’ Provide a Solution

Smart APIs — APIs infused with AI — can help enterprises and other organizations make the most of their APIs, Nargund said.

“Often there’s an event or a context awareness around that API,” he said, and the AI helps easily adapt and assist across its life cycle. “There’s an AI-assisted governance design when you’re building, there’s a context-aware behavior.“

When used in an architecture that includes AI agents, he added, “the state and context that are used for the agent workflows and the eventual API integration around that becomes very critical.” As APIs move data, “smart APIs help interpret and act on it. That’s where the smartness comes in.”

Event awareness is a big benefit of smart APIs: “You need to be able to bridge your APIs with streaming and real-time apps and AI agents. So there has to be an event-driven readiness around your smart APIs in large organizations.”

IBM’s API Connect for Hybrid Environment Governance

The first version of IBM’s API Connect was introduced 10 years ago. The latest iteration — version 12.1, released in December — aims to solve some of the issues enterprises face in the era of smart APIs.

“We are embedding AI into the entire life cycle of API Connect,” Nargund said. The platform includes a “guided wizard,” as he described it, to create AI-aware APIs. It includes DataPower Nano Gateway and API Studio.

IBM also recognizes that enterprises generally deploy in multi and hybrid environments. “Enterprises just don’t use one hyperscaler; they use multiple hyperscalers,” Nargund said. “They run on the edge, they run on prem.”

“So with so many different types of deployed infrastructure the API is running across, obviously it becomes very important for the enterprises to have a strong API management and governance and observability.”

Nargund works with IBM’s web methods hybrid integration program. “That is our premier offering for hybrid integration, but it provides multiple capabilities,” he said. “It’s available on the AWS Marketplace. It’s an AI-infused hybrid integration platform. It provides [mobile file transfer] capabilities, event gateway capabilities, API management capabilities, all together on a single AI-infused platform.”

At its core, he said, is a hybrid control plane. “Using a single control plane, you can look at multiple deployments of this usage of this API across your hybrid environment. That’s a very powerful tool, and that is something that IBM and the integration team is heavily betting on.”

Listen to the full episode to learn more about the latest version of API Connect and how more companies should approach scaling up their AI pilot projects.