Laying the Financial Data Foundations for the Age of AI

Finance departments are being pushed to deliver forward-looking guidance and near-real-time visibility, yet many still rely on systems built to produce and reconcile historical results. That mismatch is most visible during the financial close. According to APQC, the median close time is 18 days, stretching to 35 days among slower-performing organizations. Only 31% of organizations actively use AI in record-to-report processes, while another 39% remain in early adoption stages, suggesting many finance teams are still building the technology foundations needed for scalable automation.

Background: Expectations Outpace Architecture

In an April 2026 study of 197 finance executives by the Financial Education & Research Foundation, 81% cited manual processes as slowing the month-end close, 49% cited system limitations, and 38% cited data quality issues. Fifty-five percent named legacy systems as a barrier to transformation. Alex Curran, Chief Executive Officer at Aptitude Software, describes a decade-long shift: finance once reported accurate numbers; now it must explain what happened, predict what comes next, and recommend action in real time. That tension, she says, has sharpened over the past 12 to 18 months.

Development: Transaction-Level Capture and Common Language

Curran argues that finance architectures typically aggregate data before it reaches the general ledger, losing transaction detail, product, timing, and amount. Subscription models, digital transactions, multi-entity structures, and cross-border complexity have multiplied data volume while the underlying architecture was built for far less. Her prescription: capture every financial event at the transaction level, preserve lineage, and surface reconciliation exceptions immediately. Ciprian Porutiu, Senior Vice President of Strategic Initiatives at Marsh, frames the problem as an iceberg—lagging indicators above the waterline, leading indicators below. He lays out four sequencing steps: establish a source of truth, standardize taxonomy and data dictionaries, federate data domains with ownership close to the business, and treat reporting accuracy as a test of the entire chain. Both guests spoke on a two-part series on the AI in Business Podcast.

What to Watch

Curran notes that in banking, a margin shift from funding costs can take six weeks to appear in management accounts, while real-time profitability surfaces it the next morning. In insurance, a live P&L can show claims deterioration in a specific segment before the loss ratio compounds. The gap between 18-day closes and real-time insight remains wide, and the 31% AI adoption figure in record-to-report suggests the data foundation work is still the binding constraint.