Building an AI-Native Finance Function: A Pragmatic Approach

OpenAI CFO Sarah Friar recently outlined a vision for “finance functions built on zero-day closes,” emphasizing continuous reconciliation and decision-ready information. While mid-market companies may not replicate OpenAI’s entire approach, the underlying principles offer valuable guidance.

Focus on Decision Latency, Not Just Speed

The key takeaway isn’t necessarily closing books faster; it’s reducing the time between data availability and informed decisions. Research shows that organizational readiness (data quality, process alignment) is a bigger constraint than AI technology itself.

CFOs should ask: “Which critical decisions rely on financial information that’s often stale?” Prioritizing these areas yields the greatest ROI from automation.

Start with Targeted Automation

Rather than attempting wholesale transformation, identify specific workflows where AI can deliver immediate value:

Variance commentary - Automate analysis and generate insights
Audit support - Streamline data gathering and reconciliation
Contract review - Flag key terms and ensure compliance
Cash forecasting - Improve accuracy with real-time data integration

Track the impact through measurable KPIs like cycle time reduction, exception rates, and decision quality improvements.

Framework for Implementation:

Select a process owner and define clear inputs/outputs
Automate initial steps while retaining human oversight
Monitor exceptions and identify areas for further refinement
Expand gradually based on demonstrated value

The most significant gains come from addressing repetitive knowledge work that consumes skilled labor without necessarily requiring expertise at every step.