CIO Blog

Rebuilding Enterprise IT for the AI Epoch

Ronald White
December 17, 2025
Share this blog post

On the 60th episode of Enterprise AI Innovators, hosts Evan Reiser (CEO and co-founder, Abnormal AI) and Saam Motamedi (Greylock Partners) talk with Ronald White, the former CIO of Avanade. Avanade, jointly owned by Microsoft and Accenture, operates as a $4.5 billion organization with more than 60,000 employees. The scale and depth of its Microsoft partnership make it a proving ground for emerging enterprise technologies. Ronald explains how that position pushed Avanade to adopt AI early and why the next phase of value will come from embedding AI across the entire organization.

Ronald shares that Avanade’s CIO is expected to be at the forefront of Microsoft’s innovation. He describes the role as “the tip of the spear for everything that Microsoft does,” meaning evaluating new capabilities quickly and deploying them into production environments before most enterprises have even begun testing. Not every experiment succeeds, but each one provides insight. His vantage point gives him a clear read on the hype cycle, and his assessment is blunt. “I do not think we are as far as everybody thinks we are.” Many CIOs still struggle to move from small demonstrations to strategies that produce measurable business outcomes.

Even so, meaningful impact is emerging. Ronald sees the first breakthrough when enterprises allow a large language model to operate on internal data and replace legacy knowledge management systems. The moment employees can query their institutional memory conversationally, the organization changes. As he puts it, “You are changing the game right away.” Another consistent pattern is that early wins rarely come from massive transformation programs. Instead, they show up as targeted, high-value use cases. At Avanade, this included using AI to analyze RFPs and determine which opportunities the firm was most likely to win. The workflow was small but mighty, saving time and improving decisions.

A significant barrier to progress is that many employees think they are using AI, but only at the surface level. Ronald notes that meaningful adoption requires education. People must understand what AI can do once it is tuned to their data, configured with guardrails, and integrated into daily work. He encourages leaders to focus first on fluency with the interface, then build toward more sophisticated workflows. Scale is the accurate marker of progress. In Ronald’s words, CIOs should measure success by “the number of employees that use agentic technology and how many times each day they do it.”

Ronald also urges leaders to rethink how technology teams deliver new capabilities. Traditional IT cultures prioritize stability at launch. AI systems behave differently because they evolve through iteration. Ronald often reminds his teams that perfection at day one is unrealistic. As he puts it, “It will never work right.” What matters is rapid feedback and continual improvement. Organizations that cling to old release cycles will struggle to keep pace with user expectations and model advances.

On a personal level, Ronald says AI has reshaped how he approaches work. His rule is simple. “Anything you hate doing, use it.” Reviewing documents, sorting emails, or preparing evaluations can all be accelerated with AI. He also encourages people to treat AI as a collaborator. If a result seems off, respond as you would with a colleague by clarifying expectations. This mindset helps employees gain intuition and confidence.

Looking ahead, Ronald believes AI will replace specific roles that rely on commodity tasks, while most jobs will persist with greater leverage. His current reading about President James Garfield reinforces his view that innovation cycles are constants throughout history. New roles emerge as others fade. For CIOs who feel behind, he recommends starting with enterprise search powered by large language models because it delivers immediate value. He is also intrigued by the potential of quantum computing to redefine what is possible in data-intensive work.

Listen to Ronald’s episode here and read the transcript here.