AI can only work effectively if it has access to the right data and if there is confidence that it won’t do things it isn’t supposed to do. That’s not a crazy idea. Yet this combination is precisely what many organizations still lack. For the industrial sector, IFS aims to offer a solution with the announcement of IFS Intelligence and the underlying IFS Industrial Harness.
IFS is announcing IFS Intelligence and the IFS Industrial AI Harness at its own biennial event, IFS Unleashed. As already mentioned, it’s impossible to view one separately from the other. The IFS Industrial AI Harness is the foundation upon which IFS Intelligence is built. We’ll first discuss that foundation, then the actual product.
IFS Industrial AI Harness
If there’s one thing trending in AI these days, it’s the development of a system designed specifically to rein in agentic AI. We’ve been saying this for quite some time. Don’t rely on the built-in limitations and guardrails of AI models, take control of that yourself first and foremost. Only then can you begin working with AI with any degree of confidence.
With the Industrial AI Harness, IFS aims to provide such a secure foundation for AI to its own customers. These customers are often in (heavily) regulated industries (such as critical infrastructure) and must therefore comply with fairly strict regulations. Especially now that there’s so much emphasis on the negative aspects of AI, this has become even more important.
Not a product, but rather a framework
The IFS Industrial AI Harness is not a product, but rather a framework for running AI securely. You can view it as an extension and further formalization of the existing capabilities IFS offers in the areas of security, governance, safety, and trust in the company’s products. The Industrial AI Harness thus provides the context needed to accurately manage aspects such as access rights and other operational matters.
In this way, the IFS Industrial AI Harness is designed to provide a validated foundation for AI agents to operate largely autonomously. In doing so, the AI agents utilize existing identities and the associated rights, business rules, and similar elements. Ultimately, the entire system is also fully auditable. This last point is an extremely important part of the overall solution. It receives all this information from a central location: the IFS Cloud. IFS promises that nothing else is shared with external AI providers.
As we interpret it, the IFS Industrial AI Harness can serve as a foundation for all kinds of (agent-based) AI tasks, products, and services. IFS Intelligence, discussed below, is something the company built in-house, but the foundation provided by the Industrial AI Harness can also be used in combination with other tools, whether in-house or third-party. We believe this is particularly important for the industries in which IFS operates. In those sectors, custom software is used extensively and frequently, in addition to what IFS has to offer. That software must also be able to access the Industrial AI Harness.
IFS Intelligence
To illustrate how IFS envisions the use of the Industrial AI Harness, IFS has also immediately developed a product that operates on top of this formalized foundation. That product is IFS Intelligence.
Starting today, IFS Intelligence is part of the IFS portfolio as a universal, enterprise-wide license. Among other things, this means it works in combination with any compatible software for which a customer holds a license. There is no tiering of features. When you purchase IFS Intelligence, you get everything it has to offer. In addition, customers who purchase IFS Intelligence also gain access to AI Executions.
IFS Intelligence brings data together
Based on what we’ve read and heard about IFS Intelligence, it’s best viewed as a kind of data fabric. Within IFS Intelligence, organizations can bring together data from every nook and cranny of the business. This includes data from systems and processes, but also, and certainly, assets scattered throughout the organization.
It’s important to note here that IFS Intelligence isn’t necessarily a separate layer on top of an existing stack. It’s part of the operational systems, which means it understands how organizations work. It also leverages real-time data it receives from the IFS environment and third-party systems. Within IFS Intelligence, organizations can define the rules it must follow. This is in addition to the Industrial AI Harness on which IFS Intelligence is built.
Practical and goal-oriented
When we look at the practical application of IFS Intelligence among IFS customers, preventing unplanned downtime quickly emerges as a key priority. This is one of the biggest challenges facing industrial organizations. The data needed to prevent this is often already available somewhere, but bringing it all together, taking timely action, and thereby preventing downtime is a difficult step to take. That’s where IFS Intelligence is designed to help.
It’s not entirely clear to us whether IFS Intelligence is a further development of IFS Resolve, which emerged last year from Nexus Black, IFS’s incubator. IFS Intelligence has a broader scope but is designed to lead to much faster problem resolution. That was also the approach of IFS Resolve.
IFS Intelligence is transparent
One of the risks of developing a product like IFS Intelligence is that it could become too complex to be truly valuable. According to IFS, that is not the case here. The system features a single interface where users can ask their questions in plain language.
Of course, a great deal more happens behind the scenes. However, it’s important to know that IFS Intelligence isn’t a black box. It clearly demonstrates how it arrives at certain insights and answers. In other words, the AI is explainable and shows the sources on which it relies. The IFS Industrial AI Harness comes into play as soon as actual action is taken based on IFS Intelligence, and its role is to ensure that everything is executed correctly.
Employees can use IFS Intelligence in several ways. It’s possible to ask questions of IFS Cowork, but you can also delegate entire workflows to AI agents via Loops, IFS’s agent-based AI platform. If organizations need a little more guidance during the rollout, there’s Runway, a new component of IFS Intelligence. This should make it possible to speed up migrations and configurations.
IFS Intelligence and IFS Industrial AI Harness embed AI
The announcements IFS is making today at Unleashed are crucial for accelerating the adoption of AI in industrial environments. Effectively combining data to gain insights and then trusting that AI agents will do their jobs properly have been, and still are, two major challenges. With these two announcements, IFS is taking significant steps in the right direction. It has the potential to remove two major hurdles.
Organizations certainly don’t have to let AI agents operate completely autonomously right away. For many, placing that much trust all at once is a step too far. It is therefore possible to move toward autonomy step by step. The fact that IFS has been active for a long time in the sectors it targets should, incidentally, help this trust grow relatively quickly. In a world where AI is facing a crisis of confidence, to put it mildly, that’s already quite significant.
