Infor launches agentic AI that is less ‘me too’ and more ‘us first’ for medium enterprises. Will it fly?

Industry-specific software specialist Infor has announced its move into the agentic AI space with a new offering, Infor Industry AI Agents. 

The agentic solutions are based on Amazon Web Services’ (AWS) Bedrock – capitalizing on AWS’ $90 billion R&D investments last year – linked with Anthropic’s Claude Large Language Model. In turn, Infor Industry AI Agents are powered by another new product, Infor Agentic Orchestrator, built on the company’s cloud platform. It coordinates how agents interact with both Infor and non-Infor systems, data, and models.

Though late to the agentic party, the New York-headquartered company claims its micro-vertical focus, open approach, and multi-tenant architecture allow it to address the unique challenges and pain-points that mid-sized enterprises face, sector by sector, whether they are dairy producers, EV manufacturers, or textile manufacturers.

Indeed, an announcement from the company claims:

Infor Industry AI Agents are the only industry-specific, role-based AI agents built for manufacturing, distribution, and service industries, designed to work alongside teams by managing and automating critical workflows.

CEO view 

A bold statement, given how crowded the agentic market has become in the manufacturing and services sectors. Infor’s CEO Kevin Samuelson explains it thus: 

Manufacturing is hundreds of micro-verticals. And being able to do what needs to be done at that micro-vertical level, there’s no other provider that I’m aware of at that detailed level.

Infor is offering, as it were, a soup-to nuts-(and bolts) solution, appropriate given that Campbell’s Soup and a number of automotive OEMs are among its customers. Samuelson goes on: 

There are simple processes for receiving goods and billing in both cases in manufacturing, but they’re so starkly different that I don’t think any other companies are able to do that. And if you’re building a horizontal product that serves every single industry, then it’s much more difficult to go to that level of specificity.

And being privately-held (part of Koch Industries, which remains one of the US’ largest private companies) means Infor can focus on the long term, adds Samuelson – an implicit differentiator to those companies in the AI space that are driven by share price or this week’s VC valuation.

Soma Somasundaram is Infor’s President and Chief Technology Officer (CTO) – and, once upon a time, was its first employee (there are now 17,500 others). He adds:

We are comparing ourselves to the alternatives available in the market. Whether it’s SAP or Microsoft, those are our primary competitors, depending on the industry, rather than some boutique company that is only focused on building agents for manufacturers. If that’s what you mean by the agentic market being overcrowded, then yeah, we agree there’s a lot that customers could choose or get a boutique company to deliver.

But our view is that we want to deliver it out of the box. And these things must continue to evolve over time. So that’s why we built it in the multi-tenant cloud. Six months from now there will be more innovation, and again in 12 months.

In this way, Infor aims to “bring agentic AI to the masses” iteratively, he explains – meaning the mass of medium-sized enterprises that, while significant players in micro-verticals, lack large IT departments or the budgets to match.

Cloud challenge 

But there is a problem in reaching that critical mass, and it is lurking in Infor’s own statistics. While it has over 65,000 customers in 175 or more countries, only 15,000 of them are in the cloud – that’s 23%, less than one-quarter. The rest are still operating on premises, especially those which are in more tightly regulated sectors, like aerospace.

So, far from being “job done” when Infor transitioned to being a cloud-first provider, and Samuelson took over from Charles Phillips to drive sales and consolidate the modernized company, it is more a case of “job underway”, given that over 75% of users have yet to make the jump at all.

Intriguingly, Somasundaram tells me that Infor has just 3,000 customers in its multi-tenant cloud out of that 15,000 total, although Samuelson contradicts those figures in our one-to-one conversation later, saying it is a simple 65,000: 15,000 split. Whoever is correct, the core message is clear: Infor needs to move the bulk of its customers to the cloud so it can push its agentic and other AI offerings.

So, far from being old news as a market, or the accepted norm, it seems that the cloud is just getting started for many medium-sized enterprise users. The knock-on effect of that is simple: becoming an AI-first business is a long way down an uncertain path for a great many tech customers.

In the meantime, shadow-IT use of Web-trained gen AIs dominates, leading to the sort poor enterprise results identified by the recent MIT study among others.

To its credit, Infor is aware of the problem. Though trailed second at the Velocity Day event, in many ways the more significant announcement is a new cloud migration offering, Infor Leap, which addresses the most common barriers, implementation delays and budget overruns that prevent cloud adoption.

The company says that Infor Leap helps customers to “confidently invest in migrating their infrastructure to the cloud with a ‘love it or leave it’ opt-out clause”. The offer includes On-Time Delivery, and On-Budget Implementation elements, with the latter offering a fixed-fee pricing model, enabling customers to avoid scope creep and ballooning costs. Samuelson says:

Leap is in many ways the first stage in this process. The market, it’s shocking, but the vast majority of companies are still on premise. So, our view is that step one is we need to get them to the cloud. And that’s when all this good stuff is possible. Those industries where there’s a lot of interconnectivities are moving more quickly. Think of automotive where, if you’re supplier, if you’re an OEM and your suppliers haven’t moved, that’s important. But areas that can be slower are things like food and beverages where, uniquely, there’s weakness in a lot of sub-verticals.

He adds:

Company size plays a role as well. Bigger companies with large IT budgets are moving [to the cloud]. And smaller companies are moving at a rapid rate. But it’s really the $500 million to $5 billion revenue players where there is usually just as much complexity as the huge companies, but without the budgets. So, this is why Leap makes sense – because this is really the heart of the market. We think there’s the most opportunity as they have the biggest challenge: getting to the cloud to adopt new technology.

Infor is extending its message of simplification and pricing transparency to its AI offerings too, via its Velocity Suite offering: a simple, all-in package without hidden costs.

Controversies

At the event, Infor also addressed recent controversies in the AI space, such as that oft-quoted MIT report this autumn, which found an overwhelming majority of generative AI projects delivering negligible value. Infor calls it “the value void”, quoting McKinsey research suggesting that 75% of enterprise customers expect 20% productivity gains from digital transformation, but only 30% of projects meet expectations.

So, is Samuelson concerned that rising dissatisfaction within, and about, the AI sector might dampen enthusiasm for Infor’s agentic offering which, compared with some of its peers, might seem late to market? In fact, he is pragmatic and sees that growing skepticism as an advantage: an opportunity to add real, rather than illusory, value with AI. He says:

It’s interesting, but the tone has really changed. I would say 18 months ago it was, ‘What do you have quickly? We want to get this and adopt this!’ But with so much pressure on every company to deploy AI, what’s happened over the last 18 months is a lot of projects in gen AI have not worked out.

So, yeah, the tone has really changed from ‘Quick! What can you get me?’ to more, ‘OK, let’s walk through where we can add value and have real success.” Because the point is no longer just to get stuff going, it’s to get stuff going which drives that value.

My take

Hear, hear on that. In the meantime, Infor’s agentic AI take is a good one, in relative terms, but the challenge is simple: until it can get the bulk of its customers to transition off premises, AI remains just a promise in the cloud.

In many ways, therefore, Stockholm was the perfect venue to launch Infor’s agentic vision: a forward-looking city of islands within a relatively small, but ambitious economy – one that, in the words of a financial analyst at the event, is “doing OK”.

So, how to get customers out of their insular, old-town isolation, and onto the highway to the airport – and, from there, into the clouds?