Business

“AI is Moving Faster Than How Companies Actually Work”: Inside Fluency’s Bet on Enterprise Visibility

By Jessa Marie Dollesin

Finnlay Morcombe, Co-founder of Fluency, believes most companies are approaching AI in the wrong way.

“They focus on models, tools, and pilots before they can clearly describe how work actually gets done,” he says. “If you don’t have that foundation, you’re building on assumptions, and that makes it extremely difficult to deploy AI effectively or measure whether it is actually delivering ROI.”

That belief did not come from theory. While working in investment banking, Morcombe was tasked with documenting internal processes and writing step-by-step instructions for how work should be done. It was slow, repetitive, and often disconnected from reality.

At the same time, he and Co-founder Oliver Farnill were teaching themselves to code during the COVID-19 lockdowns.

What began as an attempt to automate documentation led to a more fundamental insight. Companies were not struggling to write processes down. They were struggling to understand how work actually happened in the first place.

That idea now sits at the core of Fluency, a platform built around a simple premise. Before companies can automate work, they need to understand it.

The Visibility Gap

For decades, enterprises have relied on a mix of interviews, workshops, and consultant-led process mapping to understand how their business operates.

The problem is that none of those methods reflect reality in real time.

“Most companies are working with a version of the business, not the business as it actually runs,” says Oliver Farnill.

“These approaches are biased by memory and perspective, and they go stale fast. By the time a process map is finished and circulated, the work has already changed. That creates a major issue when you try to deploy AI, because you’re effectively optimizing something that no longer exists in the same form, and you have no baseline for measuring ROI against real operational change.”

He adds that this disconnect becomes even more problematic as organizations move from experimentation to scale.

“To properly deploy AI, you need to understand not just the workflow, but how it behaves under automation, what improves, what breaks, and what impact it has on throughput, cost, and decision speed. Without that visibility, AI investment becomes difficult to evaluate in meaningful terms.”

That gap has become more pronounced as AI adoption accelerates. Companies are deploying tools into environments they do not fully understand, often without a clear baseline for how work flows across teams, or how to consistently measure return on AI investment.

The result is predictable. Investment goes up, but confidence in outcomes does not.

From Guesswork to Execution

Fluency’s approach is to provide visibility into work directly at the point of execution and build a continuous, unbiased view of how it happens across an organization.

Instead of reconstructing workflows after the fact, the platform captures them as they happen and turns them into structured data.

That shift has immediate implications for how companies approach AI.

“You make better decisions, faster,” says Morcombe.

“Companies can see where AI fits, where it doesn’t, and which processes are actually ready for automation. That changes deployment from experimentation to targeted execution.”

In practice, that means less time spent on discovery and fewer resources allocated to broad, unfocused initiatives.

It also changes how success is measured.

“Teams can prioritize the right opportunities, move faster, and measure impact against how work actually changed,” he adds.

Rethinking the Operating Model

The shift is not just technological. It challenges how enterprises have historically approached transformation.

The standard consulting operating model is increasingly unable to keep pace with the speed of AI. Most enterprise approaches still rely on static assessments and long engagement cycles, built for slower, more linear transformation efforts.

AI operates differently.

“By the time a recommendation is delivered, the work has often already changed,” says Farnill.

The implication is clear: enterprises need systems that evolve with their operations in real time, rather than models that periodically analyze and advise from a distance.

A New Foundation for the Enterprise

Fluency is building toward what it describes as an Enterprise World Model, a continuously updated representation of how an organization operates.

As more work is observed and structured, that model becomes a foundation for decision-making, automation, and eventually prediction.

It also introduces a different way of thinking about enterprise software.

Instead of deploying isolated tools, companies can operate from a shared, evolving understanding of their business.

For Morcombe and Farnill, this is where AI transformation becomes tangible.

“When you start from an objective view of operations, you spend less time guessing and more time deploying systems that can be measured in production,” says Morcombe.

“And for the first time, you can continuously transform an enterprise.”