The final day of IFS Connect North America 2025 in Nashville wasn’t just another vendor presentation—it was a gauntlet thrown down in the industrial AI space. Matt Kempson, IFS’s AI Strategy lead, and Field CTO James Elliot delivered one of the most pragmatic AI implementation roadmaps we’ve seen from an enterprise software vendor this year.
Let’s start with the uncomfortable truth Kempson laid out: the AI market is a cacophony of competing voices. Enterprise application vendors are scrambling to thread AI through every workflow, AI framework companies are pushing their latest tools, and—perhaps most disruptively—an estimated 10,000 startups are flooding inboxes with silver bullet solutions.
“I love a Connect event,” Kempson quoted a telco CTO he’d met recently. “I was here in 1995. I know how that gig plays.” The reference wasn’t lost on the audience—we’ve all lived through technology hype cycles that promised transformation but delivered complexity.
The stark reality? Most organizations are drowning in AI options while struggling to implement even basic use cases. Quick show of hands during the session revealed that 80-90% of attendees had used ChatGPT or similar tools in the past week, yet few had successfully deployed industrial AI at scale in their operations.
The IFS Approach: Embedded Intelligence Over Bolt-On Solutions
What sets IFS.ai apart isn’t just another AI platform—it’s the embedded approach. Unlike competitors who require separate AI infrastructures or million-dollar platform investments, IFS has woven AI capabilities directly into IFS Cloud, making them “simple and secure by design,” as Elliot emphasized.
The strategy addresses a critical pain point: organizations don’t want to manage multiple AI vendors, data governance frameworks, and integration headaches. They want AI that works within their existing processes without requiring armies of data scientists or consultants.
IFS currently leverages more than 20 different types of best-in-class AI within their products and services. This isn’t about flashy large language models alone—it’s about applying the right AI tool to specific industrial problems, from predictive maintenance to automated invoice processing.
The Cimcorp Success Story: Starting Small, Thinking Big
The session featured a compelling case study from Cimcorp, a financial services technology company that exemplifies IFS’s “start simple, scale smart” philosophy. Rather than launching with the highest-value AI use cases, Cimcorp began with quick expense reporting—a relatively low-impact but highly visible application.
Pekka from Cimcorp explained their rationale: “Starting gradually, that’s the key, and keeping our employees with it. If we would go directly into the deep end with AI…” The sentence trailed off, but the implication was clear—change management matters more than technical sophistication.
The genius of this approach lies in its psychology. By solving a universal pain point (expense reporting), Cimcorp created AI advocates across their organization. Employees who experienced the convenience of scanning receipts with their phones suddenly became curious about AI applications in their specific domains.
This mirrors what successful AI implementations have shown across industries: user adoption drives value, not the other way around.
ROI That Actually Adds Up
IFS’s business value team has quantified specific returns across various AI use cases, and the numbers are compelling:
- 75% increase in efficiency for creating specs and jobs
- 90% elimination of errors in invoice creation processes
- Documented ROI tracking for every AI transaction
These aren’t theoretical projections—they’re measurable outcomes from existing implementations. More importantly, IFS provides real-time ROI tracking, allowing organizations to demonstrate value to CFOs and boards with concrete data rather than hopeful projections.
The company has also launched an Industrial AI ROI Calculator, enabling prospects to model potential returns before implementation. This transparency around expected outcomes represents a maturation of AI vendor marketing beyond the “trust us, it’s transformative” messaging that dominated 2023-2024.
Perhaps the most significant announcement was IFS’s removal of traditional AI adoption barriers. The platform is now available to any IFS Cloud customer, whether running on-premises or in hybrid environments, as long as they can connect to SaaS services.
This addresses a major frustration point for industrial organizations—many operate in air-gapped or highly secured environments that previously couldn’t access cloud-based AI services. IFS’s hybrid connectivity option opens AI capabilities to manufacturing, utilities, defense, and aerospace companies that were previously excluded.
Pricing has also been restructured. Instead of platform fees that can reach seven figures, IFS prices AI capabilities as extensions to existing IFS Cloud agreements. This removes the traditional “AI tax” that has prevented many mid-market organizations from experimenting with industrial AI.
The Implementation Reality
What struck me most about the presentation was its practical focus. Rather than dwelling on AI’s transformative potential, Kempson and Elliot spent most of their time explaining exactly how to get started:
- Permission Management: Organizations control who accesses which AI capabilities
- Data Source Selection: Companies choose which data sets feed specific AI use cases
- Gradual Rollout: Start with low-risk, high-visibility applications
- Continuous Measurement: Track usage and ROI in real-time
This operational focus reflects lessons learned from early AI implementations. The technology works, but success depends on thoughtful change management and realistic expectations.
While the presentation focused on current capabilities, both speakers hinted at significant developments in agentic AI—autonomous systems that can perform complex tasks without human intervention. These announcements, expected “over the next few weeks,” suggest IFS is preparing for the next phase of industrial AI evolution.
The timing isn’t coincidental. As organizations become comfortable with assisted AI (like expense reporting automation), they’re ready for more sophisticated autonomous systems that can manage end-to-end processes.
What this means for ERP insiders
The AI integration window is closing fast. Business leaders should audit their current ERP AI capabilities and identify 3-5 quick-win use cases that can be deployed within 90 days. Focus on high-frequency, low-complexity processes like expense reporting, document processing, or basic predictive maintenance alerts. These provide measurable ROI while building organizational AI literacy. The key is starting now with proven, embedded solutions rather than waiting for perfect custom implementations that may never materialize.
Data governance becomes your secret weapon. While competitors struggle with multi-vendor AI architectures and data integration challenges, organizations with embedded AI platforms maintain competitive advantages through unified data governance and security models. Recent security breaches at AI startups have cost enterprises an average of $4.5 million per incident, according to IBM’s Cost of Data Breach Report. IT leaders should consider AI platforms that leverage existing ERP data structures rather than requiring separate data lakes or warehouses. IFS’s embedded approach, where AI capabilities inherit existing security protocols and user permissions, eliminates the complexity of managing separate AI infrastructure. Evaluate your current ERP vendor’s AI roadmap and demand clear answers about data residency, model training policies, and security inheritance. Organizations that solve AI governance early will scale faster than those managing multiple AI vendors and data pipelines.
ROI measurement separates winners from experimenters. The AI pilot purgatory is real—McKinsey research shows that 70% of AI projects never move beyond proof-of-concept stage, primarily due to inability to demonstrate clear business value. Organizations that implement robust ROI tracking from day one are 3x more likely to scale AI initiatives successfully. CFOs and business leaders must demand granular ROI tracking capabilities from AI implementations, not just high-level productivity claims. IFS’s transaction-level tracking and real-time value measurement should be the baseline expectation for any industrial AI investment. Create AI value scorecards that track specific metrics like error reduction percentages, process completion times, and user adoption rates. Use these metrics to build compelling cases for expanding AI implementations beyond initial pilot programs. The organizations that master AI ROI measurement today will have unlimited budgets for AI expansion tomorrow.