Red Hat is a leading provider of open-source software solutions, serving enterprises worldwide. As their teams explored AI opportunities across the company, they needed a strategic framework to move pilots into production while maintaining their collaborative culture and avoiding the technical debt that plagues many AI implementations.
At Canvas 25, Red Hat’s Distinguished Engineer Jan Mark Holzer shared how the open-source software leader navigated their AI transformation — moving from disconnected pilots to a scalable framework that eliminated weeks of manual work and accelerated innovation across the organization.
Challenge
Red Hat faced the reality that most organizations encounter with AI: enthusiastic experimentation without the foundational elements needed to scale. While AI pilots are easy to start, getting them into production sustainably often proves difficult.
The core challenges they identified were:
- Data chaos: Over 2,000 data pipelines, countless spreadsheets with inconsistent definitions, and teams using different data criteria
- Technical debt risk: Without proper guardrails, AI coding assistants could accelerate development but also create massive technical debt — using outdated libraries, introducing security flaws, and generating duplicated code.
- Production complexity: Moving from pilot to production required navigating security screening, privacy impact assessments, GDPR compliance across multinational operations, and stakeholder alignment — a process that historically took weeks or months.
Without a structured approach to synthesize use cases, consolidate data foundations, and streamline the path to production, Red Hat risked either stalling valuable initiatives or deploying solutions that couldn’t scale.
Solution
Red Hat built a comprehensive framework with three dedicated teams that scaled AI pilots across the company: Velocity, Combinator, and Steering Committee. A critical part of this framework was evaluating and selecting tools that met their stringent requirements. Miro was one of the tools that successfully passed this evaluation, becoming an AI-powered collaborative workspace where cross-functional teams could visualize complexity, design solutions, and accelerate execution.
Using Miro to bridge ideation and implementation, Red Hat:
- Designed their data consolidation strategy by visualizing how their systems — internally called the Dataverse — would connect together in Miro, mapping the architecture for MCP servers that would eliminate the need for SQL queries or complex API access across their data sources.
- Scoped and refined custom applications by using Miro as the collaborative space where teams — from creative to engineering — could define project requirements, refine deliverables, and align on visual direction before development began.
- Transformed customer engagement by creating custom Miro boards for every executive briefing center visit, where customers could collaborate and share feedback before, during, and after visits. Miro likewise serves as the hub for capturing customer insights and synthesizing feedback for product and marketing teams after each visit.
The visual canvas became the space where diverse teams — data architects, marketers, creative professionals, engineers, and their customers — could collaborate on complex problems and build solutions.
Results
Red Hat’s structured approach to AI adoption delivered measurable efficiency gains while maintaining their collaborative culture:
- Consolidated data pipelines from over 2,000 down to 200, creating a unified foundation for AI applications
- Accelerated privacy impact assessments that previously took weeks or months by using AI to screen requests and automatically send clarifying questions back to requesters
- Enabled technical and non-technical teams to leverage AI without learning complex prompt construction, democratizing access across the organization
The combination of structured frameworks and collaborative tools like Miro meant teams could identify duplicate efforts across the organization, synthesize use cases, and accelerate the path from concept to production.
As Jan Mark noted, “There are no new problems in our industry. There’s just people who have never met.” Miro became the space where those people connected.
The bigger picture
Red Hat’s success demonstrates that AI transformation requires both technical infrastructure and cultural enablement. An internal set of advocates provided the governance structure, while tools like Miro provided the collaborative workspace where teams could visualize, refine, and translate ideas into action.
By investing in data consolidation, building reusable frameworks, and creating spaces for cross-functional collaboration, Red Hat turned what could have been isolated experiments into scalable organizational capabilities. The executive briefing center boards became a model for customer-driven innovation, while the Dataverse became a single source of truth organization-wide.
Most critically, by focusing on real business problems first and technology second, they ensured their AI initiatives delivered genuine value that empowered teams and people as the priority.
