HPE's CFO: Making Agentic AI Work in Finance

Why It Matters

By demonstrating a concrete framework for measuring AI value and ensuring deterministic, compliant outputs, HPE shows how finance leaders can unlock rapid productivity gains while mitigating risk, setting a benchmark for enterprise AI adoption.

Key Takeaways

• HPE uses AI agents to accelerate finance processes.
• ROI assessment blends direct metrics with indirect speed and accuracy gains.
• Deterministic AI outputs are essential for financial compliance.
• Human-in-the-loop model ensures risk-averse AI deployment throughout enterprise.
• Partnerships with Deote and Nvidia enable private-cloud, secure AI infrastructure.

Summary

In a candid interview, HPE’s chief financial officer Marie Meyers explains how the company is embedding agentic AI across its 3,600‑person finance organization. She frames the discussion around a bespoke ROI framework that captures both hard‑line financial metrics and softer, indirect benefits such as speed, accuracy and fraud reduction. Meyers stresses that the true value of generative AI often emerges from these indirect gains, which can outweigh initial productivity forecasts.

The CFO breaks down the evaluation process into direct cost‑benefit analysis and an “indirect bucket” that quantifies improvements in transaction speed, error rates, and compliance risk. Drawing on a decade of experience with RPA and machine‑learning projects, she notes that AI delivers measurable outcomes far faster than traditional ERP or CRM rollouts, which typically require multi‑year horizons. A critical differentiator, she adds, is the need for deterministic outputs in finance—AI must return the same answer every time, regardless of who asks.

To meet that requirement, HPE partnered with Deote’s Zora platform—rebranded internally as “Alfred”—and co‑engineered a solution with Nvidia’s NIM technology. The private‑cloud architecture houses roughly half a million data elements, enabling secure, on‑premise processing for high‑risk functions like accounts payable, credit, and the weekly operations call. Meyers emphasizes a “human‑in‑the‑loop” approach, ensuring auditors and compliance officers retain final authority while AI accelerates routine tasks.

The conversation signals a broader shift for finance leaders: AI can be deployed at scale with rigorous governance, delivering rapid ROI while maintaining the determinism required for regulatory compliance. Companies that adopt a similar collaborative, risk‑aware model are likely to outpace competitors in both efficiency and financial accuracy.

Original Description

Making agentic AI work in finance involves addressing a core challenge: finance relies on precision and control, while AI depends on probability and constant adaptation. At HPE, the CFO's office has advanced AI from purely advisory dashboards to actively handling real financial tasks, such as accounts payable, credit, and collections. In CXOTalk Episode 914, Marie Myers, Executive Vice President and Chief Financial Officer of Hewlett Packard Enterprise, explains how her team established the governance, trust frameworks, and talent development necessary to make this shift successful.