top of page
Search

š—•š˜‚š—¶š—¹š—±, š—•š˜‚š˜†, š—¼š—æ š—„š—²š˜š—µš—¶š—»š—ø: š— š—®š—øš—¶š—»š—“ š—”š—œ š—£š—®š˜† š—¢š—³š—³ š—¶š—» š—™š—¶š—»š—®š—»š—°š—²

Writer: Benjamin
Benjamin
5 days ago
4 min read

Build, Buy, or Rethink: Making AI Pay Off in Finance


Where is AI delivering real value in finance, and where is it creating more work?

Ā 

We recently convened four finance and business leaders to discuss how they’re putting AI and automation to work. They shared what they’re building, what they’re buying, how they measure returns, and what hasn’t gone as planned.

Ā 

Across the discussion, the focus was practical: choosing the right workflows, understanding the full cost, and keeping people accountable for the results.


Ā 

Christina Ross, Founder & CEO, Cube

Ā 

Christina’s team takes a structured, bottom-up approach to AI: identify bottlenecks within teams and departments, define success, estimate potential time savings, and measure ROI after implementation.

Ā 

She distinguished rules-based automation (if X happens, always do Y) from AI, which recognizes patterns and works with ambiguity.

Ā 

She emphasized that flawed inputs produce flawed outputs. For CFOs, auditability is another critical concern: AI-driven decisions and data changes need a clear, traceable record.

Ā 

The financial case extends well beyond the initial build. Maintenance, updates, security, and compliance risks all factor into the cost. Tracking token consumption by team, rather than as one aggregate bill, helps clarify where spending delivers value.

Ā 

Christina also confirmed that her team at Cube uses its own FP&A software internally. ā€œWe don’t call it eating our own dogfood, but drinking our own champagne!ā€

Ā 

Major takeaway: Building with AI isn’t always the most cost-effective, let alone the most secure, choice. Between building, maintenance, security, and token spend, buying or offshoring may be the better choice. Run the numbers!

Ā 

Ā 

George Wells, Founder & CEO, Wells Group

Ā 

George sees AI delivering savings by eliminating redundant tools and reducing the need for entry-level hiring for repetitive administrative work. Automating tasks like data entry also frees existing finance staff to focus on strategic decisions.

Ā 

He regularly reviews how his team uses AI to assess its impact on efficiency and quality.

Ā 

For a practical starting point, George pointed to vendor contracts. Centralize agreements, renewal dates, auto-renewal terms, and price escalators, then use AI to identify opportunities to renegotiate, consolidate, or cancel before renewal. The savings are measurable, and the process builds a data foundation for broader adoption.

Ā 

His advice on token spend: measure the value, not just the bill. Higher spending can be justified when it delivers disproportionate revenue or time savings.

Ā 

Major takeaway: The CFO should own the AI roadmap for finance, with AI embedded in the workflows the team already runs. That accountability follows ownership of the controls, spend, and data.

Ā 

Ā 

Harley Pasternak, Founder & CEO, SFP

Ā 

Harley’s team puts AI to work from day one: every CPA gets a GitHub license and Claude connected to their terminal. AI powers automations across the monthly close, with people verifying accuracy.

Ā 

But automating a broken process can create more work. An attempt to automate a chaotic accounts payable inbox took more time to fix and manage than it saved.

Ā 

As the leader of a professional services firm, Harley defaults to buying proven solutions. For smaller tasks, he uses a practical test: if he can build something useful in roughly 45 minutes, with some adjustments, he’ll consider incorporating it into his workflow.

Ā 

He also urged perspective: financial automation tools like TurboTax have been around for decades. What matters is whether each solution delivers measurable savings relative to its cost.

Ā 

Major takeaway: Measuring the ROI of AI implementation has been a common challenge for many CFOs. It’s important to move quickly as technology changes, but forgetting to measure can lead to overspending on low-impact initiatives.

Ā 

Ā 

Marco Rƶnning, VP of Finance at Legora

Ā 

At Marco’s company, which provides an AI platform for legal work, every commercial agreement goes through a Deal Desk to review and approve pricing and contract terms.

Ā 

Finance and operations previously spent substantial time answering repetitive questions and guiding sales reps through approvals. Agentic AI now handles those frontline inquiries, freeing the team to focus on exceptions and complex decisions. As pricing and deal structures evolve, automating routine rules helps keep the process manageable.

Ā 

Marco highlighted two sources of ROI: time saved through internal Deal Desk automation and the hours Legora saves legal teams on research, document review, and drafting.

Ā 

But AI can’t fix a weak data foundation. For Marco’s growing company, cleaning up core accounting data remains a prerequisite to scaling automation.

Ā 

Major takeaway: Automate the work, never the judgment. Treat AI as the layer that removes the repetitive 80% of finance work so the team can spend its time on the decisions, exceptions, and controls that actually matter.

Ā 

Ā 

Finding Measurable Value

Ā 

The strongest case for AI in finance starts with a specific business problem. Identify the bottleneck, establish a baseline, and measure whether the solution improves the work after accounting for implementation, maintenance, oversight, and of course immediate and long-term costs.

Ā 

Clean data and clear accountability underpin all of it. Faster reporting and fewer manual hours only create value when the results are accurate, traceable, and useful.


Building, buying, and outsourcing each have a place. The right choice depends on the core business, existing workflow, the team’s capabilities, and the expense of keeping the solution reliable as needs change.



The event sponsors were Monk, IOmergent, and Rillet. It was organized by New York Tech CFO Group and Build Scale Grow.

Ā 

---------------------------------

Ā 

Build, Buy, or Rethink: Making AI Pay Off in Finance

Photo for "Build, Buy, or Rethink: Making AI Pay Off in Finance" by Emily Friedman.


What if the quietest person in the room is the one who changes the world?


Build Scale GrowĀ helps early-stage leaders discover opportunities while reducing risk. We clarify priorities, solve problems, and establish systems so you can stay focused on growth.


RESOURCES


  • The New York Tech CFO GroupĀ is a private community where more than 250 finance leaders share insights on planning, operations, and technology adoption.





  • Scale: Reach Your PeakĀ helps leaders learn and understand proven and practical scaling methods in just five minutes. Browse over 130 practical topics.





Ā 
Ā 
LinkedIn.webp

© Build Scale Grow, LLC

bottom of page