Hyland releases AI agent platform and vertical integrations
Can Hyland be all things to all customers?
Hyland kicked off an ambitious round of product releases yesterday that connects people, unstructured data and processes to prepare its customers to use agentic AI on its enterprise content in place.
Released were the Enterprise Context Engine, a constantly updating data model that adds structure to unstructured data and Enterprise Agent Mesh, a package of specialized agents that run document-processing workflows for corporate departments such as customer service, accounting and sales.
Planned for release later this year are AI Control Tower to manage agents and their interactions, as well as ontologies for education, healthcare and banking verticals.
Agent Lifecycle Management, also slated for release later this year, includes tools to catalog agents, certify them for production use against corporate standards, and provide oversight and governance. Also included are orchestration, task and document starter agents that enable users to customize their own agents without having to start from scratch.
Hyland CEO Jitesh Ghai said that AI is set to revolutionize the process of what he calls "human ETL," or extract, transform and load, a common automation that works on structured data.
Humans still manually perform ETL-like work on unstructured data in documents at most organizations; Ghai predicted that agentic AI would take much of the manual drudgery out of everyday work for Hyland's nearly 15,000 customers, whether it's loading unstructured data into systems of record or moving along business processes such as approvals or decisions.
"All of that in the structured data world has been automated by machines for decades," Ghai said. "If we can give structure to unstructured data, if we can give structure to a source of information that has business context in the underlying data, we can now automate all of this human ETL [too]."
For AI, it's getting later by the day
Ghai, the former chief product officer at data-management company Informatica, took over Hyland in 2024. He changed Hyland's direction and strategy in a positive way, said Deep Analysis founder Alan Pelz-Sharpe, by narrowing the product line and concentrating development on new AI tools, as well as on cloud apps and services to support them.
He recognized that customers were just not going to rip and replace ECM systems that have been in place for 20 years in some cases, Pelz-Sharpe continued. He oversaw a much-needed roadmap reset to migrate legacy customers -- many in regulated industries and on legacy platforms -- into an agentic AI world, where the AI runs on top.
Or, as Ghai puts it, "meeting our customers where they are, through a multi-cloud, hybrid, self-managed architecture."
Despite the needed strategy refresh and updated product focus, Hyland is in the same position as many of its peers in enterprise content management, digital experience and customer experience find themselves: Most customer agentic AI initiatives are still in assessment and pilot stages.
After two years of developing agentic AI systems, it's time for these technology vendors to demonstrate that large AI projects -- all requiring process change, solving data access and governance issues, and keeping agents under control -- are worth the investment and are not just hype sizzle.
"We need to see a long-term Hyland customer in a credit union, in a hospital, in whatever, saying, 'Wow, you know, we have done this, because the technology is there,'" he said. "These [projects] are not small undertakings -- they are asking a lot for the customer.
"And it doesn't have to be an agentic enterprise. I'd be happy with an agentic department or process. For all these companies -- and Hyland is no exception -- the problem isn't technology anymore, if it ever was. The problem is how to convince customers that this is worth the effort, and that it's going to work."
That said, content management vendors and their customers are more AI-ready than, for example, users of customer experience management platforms. They have what AI needs to thrive: enterprise data.
"Hyland is actually in a better position to deliver on this, because they own the content, which is the gold," Pelz-Sharpe said.
The Enterprise Context Engine and Enterprise Agent Mesh are available today. The industry ontologies are planned for release later this month. Agent Lifecycle Management and Control Tower are planned for the third quarter. These tools were all released or previewed at Hyland's CommunityLive user conference in Orlando, Florida.
Don Fluckinger is a seasoned B2B technology journalist with more than 30 years of experience, specializing in enterprise IT, digital experience and content management. As a senior news writer at Informa TechTarget he delivers award-winning analysis that helps IT and business leaders navigate complex technologies to enhance customer and employee experiences. Got a tip? Email him.
