Andrea You | Gaps in the Modern CFO Stack: The Generative AI Opportunity

By: Andrea You, Cathay Innovation

The modern CFO’s responsibilities extend beyond financial management, encompassing risk assessment, innovation and strategic planning. As companies navigate a complex web of global markets, regulatory changes and disruptive technologies, CFOs must wear multiple hats, balancing traditional duties with various strategic and operational roles.

As businesses and the CFO role have grown in complexity, so too has the CFO tech stack. Yet, despite advancements, gaps remain. Enter (you guessed it) generative AI. But hear us out — with its ability in aggregating and analyzing large amounts of data, automating manual tasks and surfacing context-aware insights, genAI naturally lends itself to filling the gaps in the modern CFO tech stack. 

Below, we share learnings on the current state of today’s finance stack and exciting genAI opportunities that empower CFOs and teams to stay ahead of the ever-changing finance landscape.  

The Evolution of the CFO Tech Stack — A Brief Recap

Traditionally, CFOs relied on spreadsheets and standalone software to manage finance-related tasks. With the advent of ERP systems in the 1990s (let’s call this Wave 1), CFOs gained integrated platforms encompassing accounting, finance and other functions. These platforms brought significant efficiency gains but were often expensive, needed advanced analytical capabilities and were primarily only available to larger enterprises (e.g., SAP, Oracle, Workday, IBM). 

Wave 2 gave rise to cloud-based solutions that granted CFOs access to real-time data, advanced analytics and were more appealing to middle-market buyers, but still required technical knowledge and were only used by few key players — think Anaplan, Adaptive Insights or Planful.

In today’s “Wave 3” solutions, user-centric design, quick implementation and automation are table stakes. But new startups have emerged that enable CFOs to capitalize on newfound genAI applications, leveraging contextual understanding to deliver more tailored results. While many of these new solutions are largely copilot tools layered onto existing solutions vs. standalone platforms, they’re rapidly enabling CFOs to 10x output while keeping costs lean.  

Myriad of CFO Tools Causes Analysis Paralysis — and Gaps Remain

Despite advancements across the finance value chain, conversations with CFOs shed light on several challenges they continue to face: 

1. Time spent normalizing, reconciling & aggregating information from disparate sources: Numerous tools span financial forecasting, tax management, treasury management, AP/AR and more, but most are point solutions that target a small piece of the value chain. Integrations are seldom and clunky, requiring significant staff for maintenance and reconciliation — creating a patchwork view of the business that may not be in real-time. Further, with recent macro shocks, teams need tools that help layer external data points into financial analyses to drive more accurate projections.   

2. Distributed teams demand better collaboration & visibility: Email and Excel have been the most widely used tools for communication and reporting — neither conducive to accuracy, visibility or efficiency. Teams are looking for cross-functional tools built for the digital-first enterprise with workflows for distributed teams and non-finance business units. Amid cost-saving initiatives, teams are becoming leaner but still expected to plug in, regardless of technical background.  

3. Changing regulatory environment poses fundamental challenges: Regulatory demands are becoming more stringent, putting pressure on CFOs to guarantee accuracy and compliance. While we’ve seen new tools focused on compliance and tax automation, most work takes place in siloes and requires manual inputs. For example, tools will surface new regulations, but compliance teams still need to make manual updates to existing processes and build new processes from scratch. 

Generative AI — Supercharging the CFO

The financial services sector has approached genAI with more apprehension vs. other industries and rightfully so — it’s highly regulated and demands near-perfection to protect against high-cost consequences. However, genAI presents an exciting opportunity to revolutionize how solutions are delivered, democratize access with more tailored solutions and unveil newfound efficiencies. 

We’ve seen a proliferation in genAI copilot solutions that target “lower-stakes” workflows farther from the core product or customer — making back-office CFO applications the perfect place to plug in.  While many finance workflows are highly nuanced and human oversight is still needed, genAI can 10x productivity by broadening access to data, research and helping automate manual, error-prone tasks. 

A few examples:  

• Supercharging financial analysts by aggregating disparate data sources, analyzing large datasets and uncovering insights with contextual, business-specific understanding.  
• Supporting robust financial modeling and forecasting by integrating more external data points. 
• Streamlining compliance workflows by automating audits and continuously monitoring and updating for new regulatory requirements to ensure adherence.  

Key Takeaways from CFOs — Building for the customer & maximizing value

Pinpointing how genAI will augment customer value, building with intent and targeting the right entry pain point is critical to finding traction quickly. Here’s what we learned from recent tools with strong product-market fit with traditional financial services players and fintechs along with conversations with CFOs on how startups can build to meet their dynamic needs with new AI capabilities: 

• Embed & support contextual understanding to capitalize on strengths.

As the quality of genAI tools relies on the context to which models are exposed, embedding directly into workflows capitalizes on its strengths. For example, Excel’s AI chatbot surfaces trends and builds model outputs directly within Excel’s interface. While users still need to prompt the tool and check results, having the AI tool embedded as an integral part of an existing workflow drives more sophisticated analysis with contextual understanding and user-friendly UX. 

IRL Startups: Runway embeds seamlessly into modeling workflows and CRMs, making tailored copilot solutions available without users having to leave the platform.

For example — VAT regulation and recovery are known for their intricacies with varying rates and rules based on country, goods and services. Finance teams must review and extract information in a manual process, that doesn’t scale, and is susceptible to human errors — with missed claim refunds costing companies hundreds of thousands of dollars. By embedding VAT reclaim tools into accounting flows, companies can automate data analysis extraction, verification and fraud detection with real-time reporting and analytics— accelerating processes, reducing errors and maximizing recovery. Further, VAT automation could be a startup’s initial wedge into a business, enabling it to expand into other financial offerings.  

• Build for interoperability: buyers look for integrations vs. a complete overhaul. 

Winning solutions plug into various parts of a business, allowing for “cross-pollination” across siloed applications. We’re seeing genAI applications embedded in tools that sit on top of CRM, ERP and workforce management systems. This enables teams to better understand where and why spend is happening. Finding the right point of integration is key — the optimal entry point is the deepest pain point and is critical to driving adoption at scale quickly. 

IRL Startups: Cobbler, Finally, and Pennylane start with accounting and AP/AR management or plug directly into widely used data, accounting or payroll systems and aggregate financial planning into a single solution.

• Build for collaboration: finance teams are increasingly cross-functional. 

Teams are leaner and more distributed with various levels of technical know-how. User-friendly solutions with collaboration as a core feature can help overcome barriers to adoption, such as low/no code interfaces and other self-serve solutions. Further, the recent macroeconomic landscape has made real-time KPI reporting critical and many companies have changed reporting processes to support more frequent updates, driving demand for more customizable solutions. 

IRL Startups: Mosaic and Basis offer drag-and-drop dashboards that allow users to create custom operating metrics and reports tailored to organizations’ needs — enabling quick plug and play solutions without having to learn a new UX.  

• Build defensibly: an exciting opportunity of vertical genAI.

The past few years saw the great bundling of horizontal tools, yielding a highly saturated market. However, numerous seemingly niche markets still require a level of focus but continue to be underserved. The most significant advantage of vertical AI solutions trained on domain-specific LLMs is that they can be applied to deep use cases in ways that horizontal software simply cannot. For users, vertical solutions deliver more sophisticated outputs that account for the nuances of the financial services industry and reduce the need for as many integrations across disparate apps.

IRL Startups: Parcha (finance-specific AI agents), Hadrius (AI-powered SEC compliance for financial services), and Basis (AI assistants for accountants) are good examples. From a platform’s perspective, having access to or owning proprietary data is vital to building defensibly, and this moat only widens as models gain a deeper understanding of customers.

Parting Words 

The CFO stack has come a long way, enabling CFOs to play a pivotal role in their company’s success. However, challenges persist, from data integration and real-time insights to forecasting accuracy. By embracing genAI, CFOs can address pain points and elevate strategic contributions, steering organizations toward a more prosperous future. 

At Cathay Innovation, we’re excited about the next generation of CFO tools that build with flexibility, collaboration and interoperability — enabling scale with dynamic finance teams and thoughtful integrations with genAI. While “one-stop solutions” are tempting, we believe the way to win is identifying the single, deepest pain point as an initial wedge— making product adoption crucial before expanding to other parts of the stack.

Building in the space? We’d love to hear from you! 

---

For more insights on the world of AI, fintech and startups, check out our previous posts: 

Key Takeaways from Money20/20: AI X Fintech? Not Yet.

Reality vs. Hype: How AI & Human Activity Combine

Industry Insights: Fintech’s Growing Role in the Healthcare Revolution