IFS Unleashed 2026: The ERP company that wants to connect the whole loop

I spent the week of October 12 in Orlando at IFS Unleashed, and what struck me most was how little of the main-stage conversation was about ERP. IFS started as an ERP company (financials, HR, manufacturing, supply chain), and that was my introduction to the company years ago: its ERP solutions sold into the AEC/construction world. Over the years, though, IFS built a strong business in enterprise asset management and field service into all sorts of industries – perhaps not quite rivaling SAP and Oracle, but a strong contender nonetheless. That is still the foundation. But the value proposition is now so much bigger: IFS wants to be the connected, AI-enabled layer that carries a workflow from the field asset or factory floor up to the management decision, then back down to the work order, the purchase order, and the technician.

In other words, IFS no longer wants to be just a system of record. It wants to connect the systems of record, including ones it doesn’t make.

The AI layer: a harness, not a chatbot

The centerpiece of Unleashed 2026 was IFS Intelligence, launched at the event, with its CoWork, Loops and Runway components. IFS calls this an “industrial AI harness”: orchestration, skills and connectors, context and memory, verification and escalation, permissions and audit, all built in.

Many of the customer stories were concrete examples of this in action. Cadillac F1, for example, just joined the F1 circuit. The team had parts that should have been serialized but weren’t, and that had since been consumed into huge assemblies across shop orders. Fixing that by hand took about two weeks per part type. IFS Intelligence is being used to unwind the transaction history, fix misclassified parts, and correct the source data to standardize across builds.

IFS CEO Mark Moffat has been appearing at AVEVA and Siemens user events, highlighting those partnerships, and they were front and center at Unleashed.

Siemens and IFS close the design-to-field loop

From a PLMish perspective, Siemens gave us the most interesting main-stage demo. A fictional gas-turbine operator spots a pattern of early part replacements in IFS. Siemens’ Intelligence Center X links those service records to product data in Teamcenter, and AI agents trace the problem to a design flaw: the original design’s thermal simulations never considered that a component would overheat every time a turbine stopped, and this one stopped periodically. The engineer reruns the simulation, including the overheating part, and designs a field repair for the turbine. The fix flows back to Teamcenter, then to IFS, to orchestrate a retrofit across the fleet.

We’re getting closer to closing a loop that PLM vendors have been working on for 20 years. But it’s early days: This demo assumed as-built and as-maintained data linked to Teamcenter, which is possible for an equipment OEM that services its own fleet — but much harder for a manufacturing plant owner-operator who never got the OEM’s PLM data. Baby steps — and an important first step.

AVEVA brings the process side

IFS’s other big PLMish partnership, with AVEVA, covers process plants and infrastructure. The joint offering, to be called CADI (Continuous Asset Decision Intelligence), will connect AVEVA CONNECT and its operations data with IFS’s asset, maintenance, and capital-planning through a shared knowledge graph. This is about plant data: P&ID topology, tag and asset mapping and not equipment design. CADI is in development and lining up lighthouse customers; look for demos early in 2027.

But it’s never easy

AVEVA’s parent, Schneider Electric, is buying Cognite, whose contextualization technology overlaps CADI’s middle layer, and on October 4 agreed to buy PTC — that puts Windchill and Creo (competing with Siemens Teamcenter and Designcenter) and ServiceMax (an IFS field-service rival) under the same roof as IFS’s AVEVA partner. We’ll have to see how that all works out — everyone at Unleashed was very careful not to answer questions about the impact of potential/possible/likely acquisitions.

And everyone now has a knowledge graph: Siemens Intelligence Center X, AVEVA CONNECT, Cognite, and IFS’s own semantic layer all provide some form of knowledge graph. The questions of who owns the cross-system ontology remains unanswered.

Finally, IFS’s (and, really, every COTS vendor’s) real competition is DIY: customers are already building their own MCP connectors for vendor APIs. DIY is perceived to be faster than waiting for a commercial release, and 100% keyed to the customer’s needs. IFS argues that a governed, supported layer beats a homegrown one, every time. I agree, but vendors can’t be everything to every customer at once, so we’ll see a lot of DIY as the vendors work through their priority lists.

TL;DR. IFS’s shift from ERP vendor to cross-system workflow enabler is fascinating. Manufacturing and asset-heavy industries need to connect design intent, operations, and business decisions across the systems they naturally use. IFS has a credible AI architecture, real early customer results, and, now, partners that fill a big gap: engineering.

But IFS also needs to serve its longtime ERP customers — and they seem happy with its AI direction, even if mystified by the gas turbine example. Customer speakers with titles like ERP Orchestrator and ERP Admin described how AI agents let them screen training records, process invoices and payments, and reconcile ledger entries faster and more accurately. My favorite story of the entire event came from an ERP Admin who described training an AI co-worker to reconcile expense statements. No one likes telling a colleague that they can’t charge XYZ to the company on a business trip — and now, no one has to. The AI agent will handle it. Those are unsexy tasks and topics, but they make our businesses work better.

Note: IFS graciously covered some of the expenses associated with my participation in the event but did not influence the content of this post in any way. The title image is of IFS CEO Mark Moffatt on stage during his keynote presentation, courtesy of IFS photographers. It’s also pasted below, since some browsers don’t show the title image: