IBM Consulting offers enterprises an easier path to agentic AI

IBM Corp.’s consulting division is stepping up its game in artificial intelligence with the launch today of a new service that can help enterprises automate key business processes with AI agents on their cloud platform of choice.

AI Integration Services is an IBM Consulting service that will, in the company’s own words, help customers build their AI agents and move from “productivity to performance” much faster than they can do alone.

In a blog post, Francesco Brenna, IBM’s vice president and senior partner of AI Integration Services, said the potential of agentic AI has gotten a lot of business leaders excited because it promises some astonishing productivity gains. Agentic AI refers to AI systems that can understand problems, work out how to solve them, and learn from the experience, all without any input from humans. They can potentially automate and accelerate many aspects of enterprise work, freeing up employees to focus on higher-value tasks or, in some cases, making some jobs unnecessary.

However, implementing agentic AI has proved to be a significant challenge, and a lot of early experiments have failed, Brenna said. He explained that most early experiments involving AI agents are focused on simple productivity improvements, or simply “bolting AI onto outdated business processes.”

As a result, these experiments often are disappointing, driving only marginal improvements in productivity, if at all, Brenna says. One of the main causes of this is that AI agents struggle to access corporate data, which is often fragmented and outdated, and sometimes even low-quality. That limits the capabilities of their AI agents.

“Almost every organization has legacy IT systems with disparate applications and complex integrations, which makes it harder to pull data from where it’s needed to support new AI-powered processes,” Brenna said.

In addition to the headaches around data, organizations need to train their employees how to understand and employ AI agents to improve their productivity. Yet many struggle to do that.

Brenna said it’s clear that AI agents alone are not enough. What’s needed, he argues, is a completely new approach that involves re-engineering business processes to create a superior user experience that can be combined with an orchestrated system of AI agents and the data that supports them.

IBM refers to these new business process models as “agentic apps.” Its says they require a “persona-based user experience” that integrates AI agents with the tools employees use on a daily basis. In addition, they require “multi-agent orchestration” capabilities to coordinate how multiple AI agents can work together to execute, monitor and control business process workflows. Finally, there’s a need for “optimized data management” to organize, prepare and make the necessary information AI agents require to automate the tasks they’re applied to.

This is where IBM says its consulting teams can make a difference, helping companies implement all of these changes and create agentic apps that deliver a return on investment.

Brenna highlighted the significant expertise of IBM’s 160,000-plus consultants, saying they have earned more than 135,000 industry credentials pertaining to generative AI, large language models and agentic AI, across industries such as financial services, public sector and telecommunications.

IBM’s AI Integration Services also bring the IBM Consulting Advantage platform to the table. According to Brenna, this is an AI-powered delivery platform that supports low-code and no-code approaches to AI agent development. Organizations can use it to design, build and deploy AI agents atop of any cloud platform, including IBM’s own cloud, Amazon Web Services, Microsoft Azure and Google Cloud.

Finally, IBM can help get customers started with a library of prebuilt agentic application templates geared toward some very specific industries and processes. They combine the essential persona-based user experience with powerful AI assistants created by IBM and partners such as Salesforce Inc., Adobe Inc., SAP SE and ServiceNow Inc., so that anyone can get started quickly and create automated workflows.

Customers can start with a template and customize it to suit their needs for various essential workflows, including customer service, source to pay, supply chains, banking operations and regulatory processes.

Holger Mueller, an analyst with Constellation Research Inc., said AI integration represents a low-hanging fruit for IBM Consulting due to the extremely repetitive nature of the work involves, as well as the amount of data architecting required to get AI agents up and running.

“IBM sees the opportunity here and is providing one of the first such offerings in a very timely fashion, as organizations start scrambling to adopt agentic AI,” the analyst said. “There is a lot of demand for this, because enterprises need help, especially when it comes to getting the necessary data together.”

Brenna said one global life sciences company has already employed its AI Integration Services to automate the laborious process of generating content for regulatory submissions. The result was a new agentic application that incorporates multiple AI agents to draft technical documentation that only has to be reviewed by humans before being submitted, according to IBM.

Image: SiliconANGLE/Dreamina