🎧 Listen to the full FEI Podcast episode »
Curran has spent more than two decades working with CFOs across banking, insurance, telecommunications, and technology, helping them solve complex finance and accounting challenges. That vantage point has shown her a consistent pattern: finance teams are being asked to deliver real-time information, prove AI’s return on investment, and manage risk, all while running on architecture built for month-end reporting, manual reconciliations, and batch processing.
“The finance teams are running AI on data that was absolutely never designed to support it,” Curran said. “Batched, aggregated, days old.” She pointed to survey data showing 87% of CFOs consider AI critical to their finance operations, yet only 14% can point to measurable returns. The gap, she said, is not accidental. It is a byproduct of asking AI to perform on top of infrastructure that was never built for real-time, provable answers.
From Steward to Copilot
Curran described a shift underway in the CFO role itself. Historically, CFOs have served as stewards of historical numbers, responsible for closing the books accurately at month end. Increasingly, they are being asked to become a copilot to the business, putting real-time information into the hands of marketing, risk, and FP&A teams within minutes rather than months. That shift is also changing the composition of finance teams, with organizations enabling accountants to take on more analytical, data science-style work.
An Architecture Problem, Not an AI Problem
Curran distinguished between AI-embedded products, where a vendor bolts AI onto an existing platform, and AI-native solutions, which are built to work with any AI technology of choice and can adapt as the landscape evolves. She described a new category she calls finance ERP: a composable layer that houses the accounting engine, sub-ledger, general ledger, fixed assets engine, and revenue recognition component, while connecting to an organization’s existing AR, AP, treasury, and tax systems. Rather than a multi-year, multi-million dollar ERP migration, Curran said these implementations can be completed in weeks or a handful of months.
The SCALE Framework
To help financial leaders assess whether their organization is ready to support AI effectively, Curran shared a five-part framework she calls SCALE:
- Structured, real-time information at the transaction level, as events happen rather than in last night’s batch.
- Continuous operations, with reconciliations and controls running in the background rather than crammed into month end.
- Auditability, meaning every number is traceable back to its source on demand.
- Layered controls, with governance built into the architecture itself rather than added afterward.
- Explainability, ensuring the same inputs produce the same outputs every time, a requirement when signing off AI-driven decisions to a board or regulator.
Most finance functions, Curran said, can answer yes to one or two of these five criteria. Very few can answer yes to all five, and all five are needed to support AI effectively.
The Question Every Finance Leader Should Ask
Curran offered a single diagnostic question for financial leaders: Right now, can you access trusted, real-time information about your business’s performance within 30 seconds to a minute? If the answer is no, she encouraged financial leaders to look at AI-native finance ERPs, regardless of whether their organization generates $100 million or $100 billion in revenue.
Curran also addressed financial leaders who may be nearing retirement and hesitant to take on another transformation initiative. She acknowledged the scar tissue many CFOs carry from past modernization programs that ran years over schedule and delivered limited benefit. Her guidance: this approach can be implemented in weeks or months, not years, and getting the foundation right now could define a CFO’s legacy, positioning finance to stop explaining the past and start shaping the future.
🎧 Listen to the full FEI Podcast episode »