4 practical AI wins for finance without friction: Use cases teams are running today

Summary (TL;DR)

• This post distills insights from our virtual event 'Finance Forward: Navigating AI & Strategy with a CFO Mindset' featuring Glenn Hopper (author and AI advisor) alongside three CFOs already putting AI to work inside their finance teams.
• Four AI use cases worth your time: AP automation with intelligent routing, Cash flow with predictive analytics, AI-accelerated variance analysis and AI for board meeting preparation
• Each use case is high-frequency, low-risk and designed to work with structured finance workflows – practical enough to pilot, proven enough to scale. Skip the hype and see where AI is actually driving ROI in finance

Why most AI projects fail – and what the successful 5% are doing differently

Enterprise AI adoption in finance is facing a credibility problem.

You’ve seen the headlines: 95% of AI projects never show ROI. Half don’t even make it past pilot. So what are the 5% doing differently?

Success rarely comes from custom LLMs or bleeding-edge models alone. What we’re seeing is finance teams making careful calls about where AI fits – and how to layer it into existing systems and workflows without turning the month-end close into a science experiment.

Rather than betting on AI-first ERPs or top-to-bottom replacements, teams are leaning into enhancements inside their existing ERP. Tools that extend structured workflows, preserve governance and respect control boundaries. Especially in areas tied to reporting, spend and compliance, the risk of letting AI run ungoverned isn’t worth it. Finance simply can’t afford hallucinations or half-answers.

Where AI is gaining traction: high-volume environments with consistent logic and a clear handoff to human review. Areas where automation amplifies accuracy and speed – without compromising oversight.

This post captures what is currently working. From real operators: three CFOs using AI in the field – David Samuels (DrFirst), Vipul Shah (FinQore) and Chad Wonderling (Zone & Co) – plus strategic input from Glenn Hopper, who’s helped build AI-first finance teams across startups and PE-backed firms.

Here’s where it’s working today – and why.

#1 Use case: Automating AP processing with intelligent routing

The opportunity: Streamline accounts payable by automating invoice entry, three-way matching and approval routing using predictable logic and structured inputs.

Why it works: This falls squarely in Glenn's "low magnitude, high frequency" sweet spot – the perfect zone for AI pilots. AP errors are fixable, auditable and have a contained impact, unlike mistakes in board reporting or strategic analysis. The high-volume, routine nature of AP processing shows the clearest ROI because computers excel at pattern recognition and rule-based processing across large datasets.

#2 Use case: Cash flow forecasting with predictive analytics

The opportunity: Move beyond static forecasts to AI-powered models that continuously learn and refine projections using real-time transaction data and pattern recognition.

What finance teams are doing now: David Samuels at DrFirst uses Adaptive Planning's AI tools for forward-looking cash management. He's candid about the current state: "It's not perfect. It's developing, but at least it gives us some foundation on a go-forward basis to improve and refine that."

This became essential when investors and lenders demanded daily cash visibility. David's team tracks where the money comes in, what form it is coming in – wire, ACH, lockbox – with twice-daily reports and defined processes for outflows, including payroll, commissions and vendor payments.

Why it works: Machine learning finds patterns humans miss – like correlations between seasonal sales cycles, payment timing behaviors and cash conversion patterns. It can identify characteristics that traditional cohort analysis overlooks, such as subtle relationships between customer demographics, contract terms and actual payment behavior.

This addresses a high-frequency, moderate-magnitude challenge where pattern recognition creates genuine value. Unlike static spreadsheet models that require manual updates, AI-driven forecasting continuously incorporates new transaction data to refine predictions. The technology excels at time series analysis – exactly what cash flow forecasting requires.

"The superpower of machine learning beyond being able to learn from data is computers are better at identifying anomalies, at making predictions, at clustering customer segments than we are." – Glenn Hopper, Author, CFO and AI strategist

#3 Use case: AI-accelerated variance analysis and reporting

The opportunity: Transform monthly reporting cycles by using AI to generate initial variance explanations, trend analysis and performance commentary that analysts can review and refine.

What finance teams are doing now: Vipul Shah at FinQore uses different AI tools for specific analytical tasks – Claude for business-focused analysis, OpenAI for technical research. His key insight: success depends on context-rich prompts and ensuring data is structured and segmented for the specific use case.

Chad Wonderling at Zone & Co focuses on turning "doers into reviewers" – moving teams from manual data compilation to intelligent oversight. AI takes on the heavy lift of analysis, freeing people to focus on interpretation, validation and strategic accuracy checks.

Why it works: This addresses high-frequency analytical tasks without the risk of strategic decision-making. AI can process large datasets quickly to identify variance drivers and performance trends, but humans retain control over interpretation and business context. It's perfect for monthly reporting cycles where the format is predictable but the analysis is time-intensive.

#4 Use case: Board preparation under investor pressure

The opportunity: Streamline board deck creation by automating performance data synthesis and initial variance explanations while maintaining human control over strategic messaging and recommendations.

What finance teams are doing now: This addresses a critical post-investment reality that Vipul Shah described from his investor background. Investment firms typically deploy teams of analysts for months to dissect every KPI, cohort and segment while building their investment thesis. Once the investment closes, all that analytical depth suddenly becomes the CFO's responsibility – but with the same lean team and pressure to hit financial targets.

Why it works: Investors expect institutional-grade insights from lean teams. AI helps bridge this gap by accelerating the analytical groundwork while preserving the human judgment that investors actually value. As Vipul Shah frames it: "Good news fast, bad news way, way faster." 

Board preparation follows predictable patterns but requires senior-level insight. AI can handle data aggregation, trend identification and variance calculations, enabling CFOs to focus on strategic interpretation and investor-specific concerns rather than manual compilation.

The bottom line: Finance doesn’t need more AI. It needs the right kind of AI.

Not every use case is worth your time. These four probably are. They all follow the same pattern:

• High frequency
• Contained risk
• Structured data
• Clear separation between automation and human oversight

That’s the blueprint we’re seeing work – especially for teams under pressure from investors, auditors or fast-changing growth plans.

If your CFO, controller or board is pushing for AI, the answer isn’t "go slow" or "go all in." It’s go narrow. Prove value in the right places – and build from there.

Want to learn more about how finance leaders are navigating this in real time?

• Watch the full Finance Forward event replay
• Read the full post: Glenn Hopper’s framework for safe AI deployment