If you came across Lucanet before 2024, you would have seen the words “simply intelligent” sitting underneath our logo. It was the principle our Lucanet founders chose when they started the company in 1999: finance software that does the hard thinking, so finance teams can focus on impactful business decisions. With our 2024 rebrand, we carried the spirit of the slogan forward in a new way, keeping the principle at the heart of everything we do.
This year, with AI taking a major leap, it feels for us like things are coming full circle: the combination of intelligence and simplicity is more relevant then ever – especially in our market.
After AI went through a valley in the hype cycle early last year, it is back on the steep ascent of the curve. Frontier model performance and capabilities are accelerating across benchmarks and new breakthroughs (deep thinking, task duration, agent orchestration, text-to-image creation) have elevated the limits of what’s possible.
As technology and its impact are starting to finally catch up with the ambition, we are getting ready to transform ourselves again.
What's at stake for the office of the CFO
Intelligence is reshaping every category of enterprise software. Inside the office of the CFO, across finance, tax and compliance, the bar it has to clear is higher than most.
CFOs can’t afford to buy demos. They buy based on the trust they put in the software’s ability to close the books, plan the year, calculate taxes, file the report, defend the number. The work is regulated, audited, and irreversible once it leaves the building. A wrong revenue figure in a published statement is a different category of problem from a wrong line in a marketing brief.
Intelligence inside finance and tax must be trustworthy on day one, traceable on day two, and defensible in front of an auditor on day three. Intelligence embedded in in the CFO tech stack will reshape finance and tax work.
The question is which providers will earn the right to do the reshaping.
We think that right has to be earned in three places.
Built different, on purpose
Lucanet’s journey into exploring the possibilities of AI started with our core product principles that we historically summarized under our slogan “simply intelligent”:
- Out-of-the-box business logic
- Second-to-none speed in implementation and thus time to value
- The most intuitive and easy user experience a finance or tax expert can imagine
While Lucanet accelerated cloud and platform transformation over recent years, these principles always remained our North Star, and every new product innovation or technology acquisition underwent the same scrutiny of whether or not it would stay true to our core principles – even before being deployed to the most critical software buyer among all: the CFO and their finance and tax teams.
For us at Lucanet, these principles are a key differentiator to the same extent that they are a commitment to our legacy, our functional roots, and the customers we serve.
Which is why we take these principles along with us on our AI transformation journey and into our approach to how we reimagine our products in the age of AI.
Architecture
A thin wrapper around a frontier model can write a convincing email. It cannot run a true automated legal consolidation. In finance and tax, 90% accuracy is 0% accuracy: there is no such thing as a mostly correct trial balance. Anything that produces numbers in our world must be deterministic where determinism is required, and intelligent where reasoning adds value. Conflating the two is how you ship a confident hallucination into an audited report.
When Lucanet built its CFO Solution Platform, we built the foundation of what is now the best possible starting point for making AI available in a secure, scalable, and reliable way. Intelligence is now a core layer of our CFO Solution Platform architecture, not just added on the side.
Our approach separates five layers:
- Data layer: extracting, transforming, and loading data from heterogeneous sources, creating a single source of truth.
- Semantic layer: adding context to the data and making it understandable by AI.
- Intelligence layer: adding reasoning, observability, guard rails and the ability to use tools.
- Deterministic layer: deterministic business logic to execute and calculate.
- Conversational layer: keeping humans in the loop and in control.
Intelligence is embedded across the platform: consuming the data, classifying the exception, drafting the narrative, surfacing anomalies. It exists in a secure and isolated environment – controlled and certified. The users are always in control, overseeing, coordinating and reviewing what work AI does: this is by design.
Evaluation
Most software development asks: does the feature work? Intelligence development has to ask: does the feature still work, on this customer's data, in this regulatory regime, this week? Our development discipline is built around continuous evaluations – replicating answers, scoring them against known-good outputs, shipping only what crosses the bar.