Finance teams ditch excel for claude AI, doubling productivity—but governance woes loom

By 2026, finance leaders at companies like Gusto and Tide have decisively moved away from traditional spreadsheet tools such as Excel and Google Sheets, embracing AI-powered platforms like Claude Code to revolutionize their workflows. This transition has transformed labor-intensive processes—such as preparing board decks that once required two weeks and five people—into streamlined operations where teams now create pull requests in GitHub against HTML files, drastically reducing time and manual reconciliation. As Jeff Cobourn from Gusto highlights, 90% of his team's daily work revolves around Claude Code, underscoring how agentic AI workflows have become central to financial operations.

The adoption of Claude Code has empowered non-technical finance professionals to engage directly with financial modeling and coding, breaking down traditional barriers between finance and engineering. Unlike earlier AI tools that merely summarized data, Claude enables users to interact dynamically with models, write to databases, and build finance-specific applications without requiring deep technical expertise. This democratization of coding within finance teams is reshaping roles and skill sets, with companies like Gusto even scoring AI fluency across their workforce and budgeting for AI token spend as they once did for headcount.

This shift to AI-driven, agentic workflows is not just a productivity upgrade but a fundamental redefinition of financial operations and decision-making. Leaders emphasize that while Claude-powered automation doubles output, it simultaneously necessitates evolved human-in-the-loop oversight to maintain operational excellence and governance. The integration of real-time AI fluency metrics and human-AI collaboration is becoming a strategic cornerstone, ensuring that automation enhances rather than replaces critical human judgment amid a trillion-dollar market transformation.

The transition away from Excel also reflects broader strategic decisions in financial technology, as organizations grapple with build-versus-buy tensions for ERP, FP&A software, and internal tools. By adopting AI-powered cloud code workflows, finance teams are not only supercharging productivity but also reshaping team efficiency ratios and software development practices. This evolution supports a vision of a next-generation ‘company Jarvis’—an AI assistant capable of synthesizing vast organizational data from meetings, Slack, and drives to provide real-time, voice-activated insights that elevate strategic financial decision-making.

Human-in-the-loop oversight remains indispensable as AI automates mechanical finance tasks, shifting the role of finance professionals from process-followers to critical thinkers who interpret and validate data. Didi Gurfinkel, CEO of Datarails, captures this transformation: 'The process-followers are going. All of them.' Yet, the thinker’s workload expands, as one analyst can now cover what previously required five, underscoring the explosion of human judgment required alongside AI automation.

The rapid deployment of AI agents in finance is outpacing the maturity of governance frameworks, creating significant challenges around auditability, accountability, and compliance. Surveys reveal that while over 70% of finance leaders prioritize speed over governance, only a small fraction have updated internal controls to reflect AI’s role, leaving 18-30% uncertain about who would be responsible for AI-driven errors. This governance lag risks undermining trust, as AI’s 'black box' nature conflicts with the need for traceable, reproducible, and defensible financial outputs.

Finance leaders emphasize that successful AI adoption hinges on embedding robust controls, transparency, and domain expertise into AI workflows from the outset rather than as afterthoughts. Frank Cirone of Snowflake stresses, 'That kind of control has to be built into the architecture, not added after the fact.' Priorities include audit-ready documentation for every AI-driven action, integration with existing systems of record, and outputs grounded in verified tax, compliance, and financial data—features that collectively build the trust and accountability essential for scaling AI responsibly.

The tension between the urgent pressure to deploy AI agents rapidly and the necessity for rigorous human-centric oversight and governance frameworks is palpable across global finance teams. Eléonore Crespo of Pigment highlights CFOs’ struggles with CEO-driven AI mandates lacking clear governance pathways, while Avalara’s Hugo Sarrazin warns that 'speed without accountability creates new forms of risk.' This dynamic underscores the critical need for finance leaders to balance productivity gains with refined budgeting, clear responsibility frameworks, and evolving human oversight to navigate a trillion-dollar market shift safely.

AI-powered tools are revolutionizing finance team structures by enabling dramatically leaner headcounts without sacrificing capability. For instance, StackIQ CEO Yana Schuster highlights building an enterprise platform with just 2.5 to 3 people instead of the traditional 20 to 30, while Jeeves CEO reports a 10x increase in transaction volume alongside a reduced workforce, boosting gross profit margins from 40% to over 80%. This leaner operational model is echoed by Ken Venner’s experience at SpaceX, where AI allowed him to achieve outcomes previously requiring 175 people with only six, underscoring AI’s role as a powerful talent multiplier that reshapes team evolution and operational efficiency.

Agentic AI workflows are automating complex, manual tasks such as document extraction, data transformation, and client communications, freeing finance professionals to focus on higher-value, strategic work. Eléonore Crespo of Pigment notes that AI tools can reclaim much of the 80% of analysts’ time spent on data processing, allowing them to concentrate on decision-making without fearing job displacement. Similarly, Ken Venner emphasizes AI’s role in removing mundane tasks, enabling humans to excel at what they do best. This shift not only multiplies individual productivity but also elevates the entire finance function toward proactive, value-creating activities.

The rise of AI in finance is redefining the concept of a '10x' finance professional, expanding what a single hire can achieve and accelerating the pace of change in required skills. As one analyst observes, capabilities that were cutting-edge just months ago are now outdated, demanding continuous adaptation. However, this also widens the performance gap between top-tier talent and others, making strategic hiring and talent development critical. Leaders like Jeff Cobourn and Rohit Divate are embracing AI-powered Claude agents to double productivity, underscoring the need for finance teams to evolve from reactive reporting to proactive decision influence with comfort in ambiguity and strategic oversight.

While AI enables leaner teams and higher productivity, successful adoption hinges on robust data foundations and thoughtful talent management to mitigate risks and maximize value. Finance leaders emphasize that AI’s effectiveness is only as good as the underlying data quality, with poor inputs leading to garbage outputs. Moreover, CFOs like Mark Kafkin of rula highlight that AI’s true value lies beyond cost savings, enabling fundamentally new work and faster, better decisions. This necessitates evolving team cultures that embrace AI as a strategic partner, balancing human-in-the-loop oversight with automation to unlock sustained operational excellence.

Legacy ERP platforms like NetSuite have become significant bottlenecks for mid-sized companies, imposing exorbitant total costs of ownership—often exceeding half a million dollars annually—and operational inefficiencies that demand large headcounts for routine tasks. This financial and functional drag forces finance leaders to confront a pivotal build-versus-buy decision as they scale beyond basic accounting tools, weighing the entrenched but costly legacy systems against emerging AI-native ERP solutions that promise greater agility and integration with AI agents like Claude. As one analysis notes, companies around the 200-employee mark must decide whether to continue layering AI tools atop legacy ERPs or pivot entirely to newer platforms designed for AI-driven workflows.

The build-versus-buy dilemma in finance technology is nuanced, hinging on organizational scale, core competencies, and the complexity of required solutions. Larger enterprises often prefer purchasing established AI-enabled FP&A platforms to avoid the substantial engineering and maintenance overhead of building internal tools, especially when durability and customer-facing capabilities are critical. Conversely, simpler rule-based workflows layered on existing databases remain feasible for in-house development. As one finance leader put it, the question boils down to whether building something akin to a Salesforce clone aligns with the company’s strategic focus or if it’s wiser to invest in specialized external solutions.

At the heart of successful AI adoption in finance lies a robust and meticulously maintained data foundation. Finance leaders increasingly recognize that without high-quality, complete, and accurate data—particularly within systems like CRM—AI tools cannot deliver meaningful outcomes, succumbing instead to the classic 'garbage in, garbage out' pitfall. This imperative drives a reassessment of foundational data layers as firms transition from legacy tools like Excel to AI agents such as Claude, underscoring that strong data architecture is the linchpin enabling agentic workflows that can double productivity and enhance human-AI collaboration.