The CFO Office Imperative for AI Investment: From Cost Savings to Value Creation

Every AI opportunity reaches the CFO's desk as a funding request, and the list keeps growing while capital stays finite. The latest episode of the iOPEX Intelligence Brief asks how the finance office should decide: where an AI investment should begin, how to tell when AI is the right tool at all, and what evidence keeps the CEO and CFO funding the roadmap after the first project.

The conversation features Manish Gundecha, Senior Finance Transformation Leader at HP, who holds an enterprise-wide value creation portfolio and has led finance transformation, finance technology, shared services and controllership. He is joined by Manohar M Atreya, Chief Operating Officer at iOPEX Technologies, who leads iOPEX's AI-led business and technology transformation and operations. Eashwar Ramakrishnan, Head of Go-to-Market at iOPEX Technologies, hosts the discussion.

The CFO's Crossroad: ROI, Capital Allocation and the Cost of Inaction

Manish opens with a reminder that CFOs always stand at a crossroad on where to invest, and AI is no exception. He frames every AI investment on a two-by-two matrix of strategic objective against time to value. Because capital is finite, the trade-offs matter. But return on investment is only half the equation. CFOs now also weigh the cost of inaction:

"What happens if you don't make the investment decision? One year from now, two years from now, do you start losing competitive advantage to your competitors?"

Manish returns to one principle throughout the episode:

‍"Data is a competitive advantage. No company on the planet can afford to ignore that critical point, and AI is all about data."

Start With the Business Pain, Not the Technology

Many enterprise processes were designed 10 to 20 years ago, and many are broken. Mano's rule for choosing what to fix first is simple: start where the business is in pain, then pick the right tool. Sometimes the right tool is AI. Often it is RPA, a process fix, or a single script.

"There is this problem, that is a nail, I have a hammer called AI, and I'll go hit the nail on its head with AI. That obviously should not be where we go about things."

He shares a field service example to show when AI does earn its place. Traditional automation, route optimization and RPA had all failed to solve the problem. Then a model, given product data and the colloquial names technicians actually use, told the technician to carry 8 parts instead of 18 on a given day. His two tests: has the simpler approach already been tried, and is the problem complex enough to need a model? Applying AI to everything, he cautions, calls for real care around token economics.

From Efficiency to Value Creation

Asked whether operating expense forecasting is the right first AI investment, Manish reframes the question:

"The conversation is not just about efficiency anymore. It's value creation."

His litmus test for CFOs: run the manual forecast and the automated forecast in parallel. When the deltas stay insignificant, the automated process takes over, and the man-hours it releases move to higher-value work. Mano adds the scaling view from his field service work: a first project may deliver only a million dollars in savings, but pushed across the wider business, the same solution can reduce 15 million in inventory, cut time spent at POS terminals, and lighten the load each technician carries.

Avoiding Over-Engineering and Scaling Beyond the Pilot

How does finance know it isn't overspending? Manish offers three rules. First, speed over accuracy: fine-tune too long and the problem snowballs. Second, the 80/20 rule: solve the 80% first, and watch the cost per user, which "should go down, not up." Third, question every new version: is the feature something the user needs, or something IT wants to build? He calls for healthy friction between solution designers and business users, because

"if a great solution gets designed and the business doesn't use it, that's money down the drain."

Mano argues that scale starts with how you choose the first project: serious business sponsorship, finance and operations engaged on the business case from day one, clear success criteria, and a problem core to the organization.

"I don't like the word pilot that much, because this project is as important as, or more important than, the 99 other projects that come along because this one succeeded."

Change Management: Turning Savings Into Sustainable Growth

What keeps leadership funding the roadmap? Manish names the evidence executives track quarter over quarter: top-line growth, market share gains, margin expansion, not just margin protection, and higher customer retention. Above all, he asks whether the data now reveals insights the business could not see before, so it moves ahead of the curve instead of playing catch-up.

For Mano, translating savings into growth is a change management story, and a bigger one than the hyper-automation wave of the mid-2010s, because of reasoning, agent autonomy and the guardrails AI demands. The change has to come top-down from business strategy, with stakeholders aligned on why it matters:

"Communicate, communicate, over-communicate, and make sure people know why this is being done."

He also refuses to compromise on guardrails and responsible AI, and sees regulation such as the EU's as a sign the market is maturing.

Looking Ahead

The episode closes on a clear principle for finance leaders and enterprise leaders alike: the question is not simply where AI can be used. It is where AI can create enough strategic and measurable value to justify the investment.