Pythian’s AI Model Via Google Gemini Drives ‘Million-Dollar Outcomes’ For Customers

‘We rolled out Gemini Enterprise across our 500-person company in 27 countries and used real work—not demos—to test what scales,’ says Pythian CTO Paul Lewis.

Google Cloud all-star partner Pythian reinvented itself by rolling out Gemini Enterprise across its 500-person company, which led to the creation of a new AI operating model that is driving “million-dollar outcomes” for customers.

“We rolled out Gemini Enterprise across our 500-person company in 27 countries and used real work—not demos—to test what scales,” Pythian’s CTO Paul Lewis told CRN.

“The results were a 3X increase in active user engagement and an 80 percent reduction in mean time to resolution for database incidents across about 15,000 monthly database tickets,” he said.

Leveraging Google Gemini Enterprise, the Ottawa, Ontario-based company engineered the Pythian AI Operating Model—a new end-to-end framework designed to take enterprise AI from high-level strategy all the way into sustained production.

“Enterprise AI value does not come from making a tool broadly available. It comes from building the operating model around it: a measurable business agenda, a production-grade platform connected to the enterprise data estate, separate ownership for adoption and transformation,” Lewis said. “Also, an entirely new skill set to manage observability, drift, and the lifecycle after launch.”

How Pythian Is Driving AI ROI And Outcomes For Customers

Pythian, a Google Cloud premier partner, said its AI operating model “consistently unlocks million-dollar outcomes” for customers.

For example, one customer who specializes in knowledge management leveraged Pythian’s AI operating model, which was able to automate 10 percent of the client’s 20,000 annual IT tickets into “no-touch” resolutions.

This saved the client over a million operational hours, Lewis said.

Another use case was for a customer that specializes in supply chains. Pythian built custom agentic supply chain tools on Google Gemini Enterprise, which compressed forecast-matching cycles from several weeks to two or three weeks across 70 global manufacturing sites.

“The workflow behind that result is deliberately practical: an agent reads and enriches tickets, searches the relevant knowledge, and generates a mini runbook before an engineer starts work,” Lewis said.

“That changes the economics of operations because it improves the workflow end to end, rather than saving a few minutes in isolated tasks,” he said.

Common AI Use Cases In 2026

Lewis said Pythian customers are less interested in another disconnected pilot.

“Customers are more interested in identifying the workflows worth reimagining,” he said. “Or connecting AI safely to CRM, ERP, database, and knowledge environments.”

Businesses are seeking to keep AI agents accurate and reliable once they are in production.

“The question is shifting from ‘What can AI do?’ to ‘Which workflow should we change first, what outcome will prove value, and who will operate it once it is live?’” Lewis said.

How Pythian’s AI Operating Model Works

The Pythian AI operating model framework consolidates strategy, execution, and operations into a continuous loop of four key pillars—starting with field CTO strategy and governance.

Pythian’s field CTO practice provides executive advisory to establish steering committees and clear value metrics. The team audits operations to build a prioritized backlog of high-ROI use cases before development starts.

Next, Pythian establishes a secure foundation on platforms like Gemini Enterprise and connects AI directly into CRMs, ERPs, and database estates to ground models in real corporate context.

This is followed by Pythian’s dual center of excellence (COE), which handles adoption and change management, as well as engineering custom-coded AI agents and complex agentic workflows that integrate into core data platforms for autonomous operations.

The fourth pillar is Pythian’s AI production management practice—dubbed XOps—that provides continuous monitoring, prompt tuning, and model observability needed to keep agents performing without breaking core workflows.

“That is the story behind Pythian’s AI operating model. It is not a claim that one technology solves every problem, it is a repeatable way to move from strategy and prioritization through delivery and into sustained production value,” Lewis said.