One Agent, Every System: What We Took to InTouch in Lucca

The question worth asking in finance technology right now is not whether AI belongs in the finance function. It’s how to make AI useful across all the systems a finance team already runs, rather than confining it to any single one.

That was the question we brought to Tagetik InTouch in Lucca this year. More than a thousand people, three days in Tuscany, and the first time we put our new VantagePoint AI Agent in front of customers and prospects.

"With AI you will do this." Not "AI will do this."

AI was the dominant theme, as it is at every event this year. But this time, the framing was different. Tagetik didn't present AI as a set of features. They presented it as a delivery function. The distinction comes down to a single word: "AI will do this" versus "with AI you will do this." It becomes about what your team can do with it.

That matches what we hear constantly. One of the most common things clients raise with us is a version of the same problem: they know they need to do something with AI, but they don't know where to start. The mandate is there but most often the first step isn't.

The problem with AI as an island

The reason so many teams are stuck is structural. Many software companies are still building AI within its own product boundaries.

"The challenge with AI is that technology companies are developing it as an island," says John Fuggles, Partnership Manager at VantagePoint. " It sits within their product. If you have five pieces of technology in your finance function, you have five AI-embedded functionalities. Then you invest in an enterprise solution on top, and now you have another AI agent. " The result is fragmentation in the place that needs to be joined up.

A middle layer, not another silo

This is broadly where the rest of the market is heading.

"For an organization of our size, we're slightly ahead of the curve," says Sid Burrows, EMEA Director at VantagePoint. "What we've built lets clients engage with real functionality inside software they've already purchased, but through an agentic layer. And rather than limiting it to a single platform, we want to be the middle layer between clients and their technology of choice."

Salesforce offers the clearest reference point, having opened up most of its functionality to an agentic layer. The principle is the same: give clients the infrastructure to work with their technology through AI, rather than navigating five separate tools and five separate interfaces.

Sid sees this as the question that will define the next three to five years. Vendors will take one of two paths. Some will build walls around their in-platform AI to keep customers locked in. Others will open their functionality to the major AI players and retain customers by letting them work through whatever language model they already use.  

It is not all about AI

For all the focus on AI, one other theme came up repeatedly: reliable data, delivered quickly. Teams want to see how any part of the business is performing in the moment, and to trust that the numbers are correct.

The value is not the data itself, it is trusted data delivered quickly enough to change decisions. Access it in close to real time and you can move before the next close consumes the month. As Sid puts it, most teams spend so long assembling and reviewing data that they have barely acted on it before the next reporting cycle begins. Less time spent wrangling numbers, more time spent doing something with them. Teams have wanted this for years, and most still don't have it readily available.

Enterprise AI is harder than it looks

A measure of realism came from the delivery side. Individuals are more productive with AI now, and almost everyone feels it. But deploying AI across an entire organisation is a far larger undertaking than most expected.

"Many people using AI are seeing productivity gains, particularly in analysis and repetitive work," says VantagePoint Principal Consultant and CCH Tagetik Lead Nik Presern. "But deploying it at an enterprise level is a much bigger beast than everyone was expecting."

Nik also expects the market to consolidate. The finance software landscape keeps adding more tools, with new releases landing almost monthly, and he thinks the next few years will thin that field out through mergers and acquisitions. Beyond that, he sees attention shifting toward faster integration, audit and workflow tooling, and eventually clients using these platforms to build their own solutions on top.

The case for being in the room

One observation ran against the AI current entirely. The more every conference sounds the same, uniformly AI-led, the more the human side of these events matters.

inTouch understood this. Much of it was built around people spending time together, and that is what people valued most. The conversations over coffee, something learned from a peer over a drink that no slide would have delivered. As more of the work becomes automated, the appetite for being in the same room as another person only grows.

Where this leaves us

The response in Lucca was strong and we have several commercial conversations underway. If you are already invested in Tagetik, an AI capability that reaches beyond the platform without asking you to abandon what you have already built is a compelling next step, provided it is governed properly and integrated around the systems and controls finance already relies on. We are not here to replace the systems you rely on. We are here to connect them, and to give your team one intelligent, and governed, place to work from.