Disconnected and Disparate Data
Before Datarails, La Fosse’s finance team operated across disconnected systems with no easy way to see the full picture. Financial data, CRM activity, and people metrics all lived in separate places, making it difficult to get a consistent, reliable view of performance.
As Urvesh puts it, there were “hundreds of versions of the truth.” In practice, the team spent more time reconciling numbers than using them.
The impact extended beyond finance. Without a connected view of their data, key insights were difficult to access across functions, limiting how effectively teams could use that data in day-to-day decisions.
Even straightforward questions became manual, time-consuming exercises. If Urvesh wanted to understand why department costs were up, someone had to pull the numbers, investigate the context, and piece together an answer from multiple sources. It took hours, interrupted other work, and still didn’t always provide a clear, confident explanation.
For a CFO who needed fast, reliable answers to run the business, that wasn’t good enough.
Where Finance Meets AI
Datarails stood out early for its ability to fit into the way La Fosse already worked, without adding complexity.
Its Excel-native approach meant minimal change, while its ease of use allowed the finance team to own and maintain the system without relying on external specialists. At the same time, its flexibility made it easy to connect multiple systems into one environment.
Datarails being Excel-native was big. Simple to implement, flexible connectivity, and future-proofed, that’s what made the decision easy.
Urvesh PatelCFO
La Fosse started with a NetSuite integration, but didn’t stop at financial data. The team integrated their CRM and is preparing to bring their HR system into Datarails, creating a single, connected view of the business where any question, whether financial, operational, or people-related, can be answered from a unified data foundation.
We’ve plugged in our CRM, and we’re about to plug in our HR tool. With that, we can get any information out of the system.
That foundation set the stage for the next step: the introduction of the FinanceOS AI Connector, which enables AI tools like Claude to work directly with live, structured data across the business. Instead of relying on static reports or re-uploading data, answers reflect a consistent, up-to-date view.
Within a single setup session, Urvesh and his IT lead connected their data and defined how key tables and fields should be understood.
We were working within 15 minutes. It learned the system, understood our Datarails tables, and then I just asked it some finance questions, and it got them right.
From that point on, Urvesh was able to query his data through Claude and get immediate, reliable answers, grounded in the Datarails FinanceOS operational layer.
Complete Answers in 10 Seconds
The impact was clear, even in the first few weeks.
In one early example, Urvesh wanted to understand why marketing costs had increased. Instead of pulling data manually, he turned to Claude and turned on the FinanceOS connection. Within seconds, it returned a full explanation, combining a clear executive summary with a detailed, itemized breakdown of the underlying drivers.
It gave me more detail than anyone else could have given me, plus an executive summary. That would’ve taken someone two hours. It took 10 seconds.
Because the answer was grounded directly in FinanceOS’ connected financial and operational data, it wasn’t just faster, it was more complete and immediately usable.
The team quickly expanded into more advanced use cases. For quarterly reporting, Claude was used to generate a full quarterly business review deck using the live data pulled through the FinanceOS AI Connector, including structured narrative and key insights, based on existing Datarails outputs.
It produced a 45-page deck, nicely formatted, with a good layout, and proper storytelling. That would’ve taken over a week.
Beyond reporting, La Fosse began applying the same approach to forward-looking analysis.
One example was attrition risk. In a business where it takes nine to twelve months to fully train a consultant, losing employees has a significant impact. Urvesh analyzed CRM activity trends such as BD calls, meetings, and pipeline creation to identify patterns in performance.
Using that connected data through the FinanceOS AI Connector, Claude built a multi-factor model that surfaced early signals and activity patterns.
It came up with something way cleverer than I could have. It built a scoring model for attrition risk, and when we cross-referenced it against recent departures, three of the people it flagged had already left. If we’d had this earlier, we would’ve known.
What started as individual use cases is quickly becoming a more consistent, structured way of working with data and AI.
Before, everyone was using AI tools on the side, to do their own bits and pieces. Now it can become a finance analyst, a digital worker that sits with you and does actual work, but also drives higher-quality output.
A New Era for Finance
La Fosse is still early in its FinanceOS journey, but the direction is already clear. The focus is shifting from experimentation to building a structured, repeatable way of working with AI across the business.
The team is rolling out access more broadly and investing in training to ensure consistent, effective usage, while prioritizing use cases like automated reporting, anomaly detection, reconciliations, and internal controls, areas that were previously too time-intensive to tackle.
There’s stuff we’re simply not doing right now because we don’t have the time or capacity. This will let us do it.
For Urvesh, the impact goes beyond efficiency. It fundamentally changes the kind of work the team can focus on.
It’s not just efficiency, it’s higher-quality, world-class output. Now it means people can do much richer work.
Urvesh PatelCFO
