The practice behind
The AI CFO
The intersection of AI and financial planning for small and mid-sized businesses is one of the most underserved spaces in B2B finance. No individual thought leader, Big 4 firm, or SaaS vendor currently owns it with authority. This practice exists to fill that gap.
The practitioner perspective
After more than a decade working at the intersection of strategic finance and high-growth operating environments, one pattern became clear: the tools and frameworks that define enterprise finance — sophisticated forecasting models, AI-driven analytics, driver-based planning — are almost entirely inaccessible to the companies that need them most.
Small and mid-sized businesses with $1M–$100M in revenue represent the vast majority of the economy. They face the same fundamental FP&A challenges as large enterprises — forecast accuracy, close cycle inefficiency, scenario modeling gaps, board reporting overhead — but with a fraction of the budget, headcount, and institutional knowledge.
At the same time, AI has materially changed what is possible in finance. Predictive forecasting now shows 30–40% accuracy improvements versus traditional methods. Automated reconciliation can compress a 15-day close to 5 days. Natural language querying lets a CEO ask financial questions without a finance analyst in the room.
The problem is that the guidance available to SMBs on AI in finance is almost entirely produced by vendors with tools to sell, or by enterprise-focused advisors serving Fortune 500 clients. Independent, practitioner-led, vendor-neutral guidance is essentially absent.
This practice exists to provide it. The content is free. The advisory is priced for the SMB reality, not the enterprise budget.
Three beliefs that shape every engagement
Honest about AI limits
AI amplifies whatever data quality and process design you start with. A tool that costs $50,000 per year produces garbage outputs on garbage data. Most AI finance projects fail because of this — not because the AI does not work. I tell clients what AI can and cannot do before they buy anything.
Vendor-neutral by design
Datarails, Cube, Mosaic, and Jirav are all credible tools for specific use cases. None of them can credibly recommend each other. I have no financial relationship with any vendor. My recommendations are determined by your actual situation, not by partnership agreements or referral fees.
Data before tools
88% of spreadsheets contain errors. 49% of CFOs report being blocked from decisions by poor data quality. The most expensive mistake in AI finance implementation is buying a sophisticated tool before the data is ready to support it. The data readiness work is unglamorous, undervalued, and more important than any tool selection.
The AI FP&A Maturity Model for SMBs
I built the first AI FP&A maturity model designed specifically for companies under $100M in revenue. Six levels, each with documented prerequisites, realistic ROI profiles, and honest risk assessments. No enterprise assumptions. No vendor bias.
Read the maturity model →What makes this different
Six positions on AI in finance that the vendor community will not say — each backed by evidence.
"Most SMBs need less AI, not more — they need clean data first."
80% of AI finance project failures stem from poor data quality. Organizations treating data as a product are 7x more likely to deploy AI at scale.
"Your $50K/year FP&A tool is only as good as your $0 data governance strategy."
The bottleneck is almost never the software. It is the inconsistent chart of accounts, the missing historical data, and the process no one has documented.
"AI-generated financial analysis is systematically biased — and the industry is not talking about it."
Harvard Business School research found that LLMs prefer technology stocks and large-cap investments, exhibiting confirmation bias that persists under counter-evidence.
"The fractional CFO model is broken without AI — and AI without a fractional CFO is useless."
Both are necessary complements, not substitutes. The tool without the strategic judgment produces fast, confident, wrong outputs. The strategist without the tools spends 45% of their time on data collection instead of analysis.
"The biggest risk in AI finance is automating chaos."
RAND Corporation documents that automating poorly understood processes is the most common AI implementation failure mode. Speed without clarity makes the problem worse.
"Human expertise in finance is underrated, not overrated."
"AI is overrated because humans are underrated. People do not recognize how versatile, talented, multifaceted human capabilities are." — Nobel laureate economist Daron Acemoglu.
Work with someone who will tell you the truth.
Independent. Vendor-neutral. Built on practitioner experience, not vendor demos. Start with a Readiness Audit.
View Services