IFS Loops Moves Industrial AI From Digital Workers Toward Enterprise Execution

Key Takeaways

⇨
Digital Workers handle defined jobs, but the larger process improves only when tasks connect across systems.

⇨
Governance, recovery, and clear accountability decide whether agents can be trusted in daily operations.

⇨
IFS now needs repeatable results across plants, assets, and systems, beyond individual customer wins.

Industrial AI is moving beyond helping employees make decisions to completing work across the business. That is an important step for manufacturers, service organizations, and asset-intensive companies. But an agent completing one task does not mean the entire process works better. The opportunity is to connect those tasks, the people involved and the systems that keep operations moving.

Following IFS Unleashed 2026, I sat down with Somya Kapoor, CEO of IFS Loops, on KramerTALK to discuss how Digital Workers are changing industrial operations and where she sees agentic AI headed. Her vision moves beyond individual agents toward coordinated execution across business processes. The challenge is getting customers from individual successes to broader adoption while delivering measurable results.

That challenge connects directly to my Forbes analysis of the ERP adoption gap, where I examined why software innovation often moves faster than customers can change how they operate. In industrial environments, the consequences are more immediate. A missed part, delayed service call or production interruption can affect the entire operation.

IFS’s announcements at Unleashed reflect how the company is approaching that challenge. The introduction of IFS Intelligence brings together Cowork for assisted work, Loops Digital Workers, and Runway for software development and configuration. Its Industrial AI Harness is designed to manage permissions, verification, escalation, and audit before agents act on systems of record. Together, these capabilities move IFS toward coordinating work across industrial processes rather than automating individual tasks. Whether customers can adopt and scale that approach is the bigger question.

IFS also discussed Continuous Asset Decision Intelligence with AVEVA, building on a partnership first announced in May 2026. The idea is to connect engineering and asset data with maintenance decisions and the work that follows. For a utility, an equipment anomaly could lead to a repair, deferral, or replacement decision and the related work orders. This is where industrial AI could make a difference, provided the integration work and customer results support the vision.

Digital Workers Bring Industrial AI Into Daily Operations

IFS Loops gives Digital Workers defined jobs, from supplier-order processing to turning service requests into work orders. For industrial customers, that is a better place to start than deploying agents without a specific operating problem. The question is what happens to the larger process when those tasks are completed. Kapoor described the problem, “It was actually things taking too long to get executed.” She said the focus was on efficiency and growth, “not about labor replacement.” That distinction should guide how customers measure success.

Kodiak Gas Services is one example worth watching. Its Material Replenisher is in production, helping field technicians find and source parts across about 150 warehouses. Kodiak has seen an 87% efficiency gain in the replenishment process, a measured result from live use rather than a projection. The bigger question is how that gain holds up as more of its technicians use the tool day to day.

Kapoor told me how Kodiak used insights from the Digital Worker’s findings to improve parts data by consolidating duplicate entries and standardizing labels. Beyond helping technicians find parts, it gave Kodiak a better understanding of its inventory. In addition, Kitron is addressing supplier coordination and sourcing delays, while KLN Family Brands is processing thousands of purchase orders a month with a 60% efficiency gain. The value goes beyond faster transactions to fewer disruptions and more reliable customer commitments. Kapoor also urged customers not to “wait to boil the ocean.” I agree. Kodiak shows how starting small can improve data and processes, building the trust needed to give Digital Workers more responsibility.

Connecting Individual Tasks to Industrial Execution

In my earlier ERP Today analysis of IFS Loops, I examined how agents move from providing insight to taking action. Industrial execution goes a step further. Finding duplicate records differs from ordering a part, and ordering a part differs from making sure it arrives where it is needed. Inventory may appear available even when it is allocated, on quality hold, or needed for a higher-priority customer.

My InfoWorld research on trusted context makes the same point from a data perspective. Access does not tell an agent what the business needs to do next. I look at this through decision architecture, connecting context, decision logic, authority, orchestration, and execution. For IFS, the challenge is keeping that chain intact as Digital Workers move between procurement, inventory, maintenance, and finance.

IFS says its Industrial AI Harness will manage orchestration, permissions, verification, escalation, and audit. Those capabilities matter most when work crosses systems or something goes wrong. If a purchasing agent creates an order but the inventory or finance update fails, can the system identify the incomplete transaction, recover or reverse it, and alert the right person? A demo shows the task. Recovery shows whether the workflow can be trusted in daily operations.

Unleashed also highlighted Eight Group’s early use of Nexus Black Resolve in service operations. IFS says the company is moving toward IFS Cloud and Loops Digital Workers, but it still needs to measure the expansion. Customer stories become more useful when they show the starting point, deployment effort, exception rates, and whether improvements hold up after go-live.

The Competition Is Moving Toward Connected Execution

IFS is not alone in moving AI closer to business execution. Microsoft is connecting Dynamics 365, Copilot Studio, Azure, and Fabric. SAP is advancing Joule and business-data integration, while Oracle is linking Fusion applications with governed agent execution. The approaches differ, but customers face a similar problem. The work rarely stays inside one application or one vendor’s system.

In my InfoWorld analysis of Microsoft’s data strategy, I argued that connecting more data is not the outcome. Revenue, productivity, and faster decisions are better measures. SAP customers are at different stages of ERP modernization. Oracle’s separation of AI reasoning from authorized business actions is worth watching, but the test remains how it works across systems and who stays accountable.

Industrial vendors are taking their own routes. My Forbes analysis of QAD | Redzone examined its federated intelligence approach to connecting ERP, plant operations, quality, and workforce knowledge. Its expanded manufacturing intelligence direction was still targeted for 2027, so customer results remain to be seen. IFS has depth in asset management, service, and manufacturing. Its opportunity is to connect work across those functions without asking customers to take on another integration project. I do not expect one vendor to own every part of enterprise execution.

Digital Workers Change Roles, Not Just Workflows

Kapoor estimated that technology accounts for about 40% of the deployment challenge. She compared Digital Workers to interns who need clear responsibilities, process knowledge, and supervision before being given more work. I like that comparison because it recognizes the role employees still play. Someone who once entered orders may instead resolve exceptions, monitor Digital Workers, and make supplier decisions. That is a change in how the business operates, not an automatic productivity gain. As she described it, employees move toward “looking at exceptions” and “managing agents.” That requires training and clearer accountability, not just time returned to a team.

My Forbes work on Industry 5.0 and human-centric operations makes a similar point about the workforce. One manufacturer may use time saved to increase production or improve service. Another may change employee roles. Both need leadership, training, and accountability. Technology enables transformation. People determine the outcome.

Go-live is only the beginning. Support teams know where the same problems keep coming back, partners understand customer configurations, and customer success teams should connect those lessons to process improvement. IFS can use experience from Kodiak, Kitron, and KLN to help customers understand what data, workforce preparation, and operating changes are needed for the next workflow. Sometimes fixing parts data or changing a process matters more than adding another Digital Worker. Customer success starts with advocacy and is measured by outcomes.

Digital Crews Could Change How Industrial Work Is Coordinated

Kapoor expects Digital Workers to evolve into Digital Crews that coordinate larger workflows, potentially within a year. That is her outlook, not a promised customer rollout. This is an important distinction. One agent might successfully place an order while procurement, inventory, and finance still fail to produce the right overall result. Digital Crews will need to manage the handoffs, exceptions, and people involved, not simply pass tasks from one agent to another.

Customers can begin with a specific operating problem, map the people and systems involved, and measure what improves once the work is in production. Processing a purchase order faster is useful. Keeping a production line running because the correct part arrives on time is a better result. That is the standard for expanding Digital Workers into more connected workflows. Kapoor cited customer reports of “20 hours per week per person” saved on certain activities and examples of Digital Workers handling much of purchase-order processing. Those are interview examples, not independently verified benchmarks. The business still has to show what improved with the time returned.

IFS’s Next Test Is Repeatable Industrial Results

IFS has production examples and a direction for connecting Digital Workers to industrial workflows. Next, I want to see how consistently customers can reproduce those results across different plants, assets, and systems. IFS should separate measured gains from adoption estimates, show what happens when cross-system work fails, and document what customers need to move beyond the first task. Microsoft, SAP, Oracle, and industrial competitors face the same test.

ERP is becoming the execution layer, but industrial execution is more than completing individual transactions. The opportunity for IFS is to connect Digital Workers, employees, and operating systems so the business runs better from one decision to the next. Customers will judge that progress through fewer interruptions, stronger service, and sustained results, not the number of agents deployed. AI value does not arrive alongside ERP. It arrives through it.