Chicago by Saviynt: From Skepticism to a Real Value Demonstration of Agentic AI
When it comes to artificial intelligence, I consider myself more of a skeptic. Not skeptical in the sense that I doubt its existence, that would be fake skepticism. Rather, I am skeptical because the term has been widely overused.
In recent years, what we have mostly seen is a lot of marketing and communications around so called “AI” that, in reality, looked much more like advanced automation or simple conditional logic.
In other words, a lot of noise and very little real intelligence.
So it was with that critical mindset that I explored Saviynt’s new Chicago release, available in General Availability for only a few days now. And, quite unexpectedly, I was genuinely surprised. Not by a cosmetic demo or yet another gimmick, but by a feature that finally shows what AI can truly bring to the world of Identity and Access Management, or IGA.
The feature is called Integration and Onboarding Agents.
What are we actually talking about? A strict and literal application of agentic AI. Not a text generator, not a chatbot, not a recommendation tool. An autonomous agent designed to execute real operational tasks.
Its objective is clear: automate the management of applications that have historically been considered “disconnected.” Applications with no APIs, no connectors, the kind that have forced teams to rely on manual interventions that are time consuming and rarely reliable.
The agent acts like a true virtual service desk. It interacts directly with application web interfaces, currently limited to web applications, to create, modify, disable, or delete user accounts. All of this is done in a traceable, standardized, and governed way.
And that is where the difference is fundamental.
We are no longer talking about content generation, but about execution. The agent is autonomous, operates in place of a human, and applies governance rules consistently, even across systems that were previously difficult, or outright impossible, to properly integrate into an IGA program.
The platform administrator simply has to converse with SAVI, Saviynt’s chatbot that allows natural language interaction, to explain how to connect to and manage access for an application. SAVI then converts that information into something reusable at scale.
Let’s be clear. If the promise holds at scale, we may be witnessing the end of disconnected applications. May they rest in peace.
At Wepoint, this type of innovation does not leave us indifferent. It speaks directly to our core areas of expertise: artificial intelligence, cybersecurity, data, information technology, and enterprise architecture. It is precisely at the intersection of these domains that AI truly makes sense, when it finally solves concrete, long standing, well known problems.
With Chicago, Saviynt is addressing a real and persistent gap in Identity and Access Management projects. Now the question is how far the vendor will push the boundaries of IGA. The next release, Delhi, announced for 2026, should provide a clearer signal of the maturity and true trajectory of this agentic AI.
So yes, the skepticism is still there. But it is being seriously put to the test.