The AI Engagement Lifecycle with Endava, Dava.X.AI, and Miro
Agile and Project Management - DrunkenPM RadioShow Notes
Endava set out to reinvent how IT services get delivered in an AI-native world. Dava.X.AI is the group inside Endava driving that shift. Miro is the workspace they're building it on. Together they've shipped Davaflow — an AI engagement lifecycle that runs from signal to explore to govern to evolve, with humans and agents working alongside each other the whole way. On the latest drunkenPM Radio, Matt Cloke (CTO, Endava), Joe Dunleavy (Global Head of Dava.X.AI), and Dave Ross (Evangelist at Miro) talk about how the three came together, what real partnership looks like when two roadmaps quietly merge into one journey, and what it took to actually make it work. Full episode is live.
Takeaways - Reframe from value capture to value creation — don't use AI to make existing processes 10–15% better, reimagine the work entirely from an AI-first perspective. - IT services hit its "ChatGPT moment" two years ago when clients and analysts started asking Endava why they were needed if AI could just write the code. - AI is becoming a method, not a tool — just as Agile organized humans to do work, this new method organizes agents and the people working alongside them. - The technology question is largely settled; the change management question isn't — quality has hit the inflection point, and what remains is the human shift, much like Waterfall-to-Agile. - DavaFlow runs in four phases — Signal, Explore, Govern, and Evolve — moving from identifying the right problem, through virtual personas and prototypes, to governing agent output, to evolving in production. - Chat is the wrong interface for most of this work — humans are visual creatures, and tools like Miro give LLMs context that a chat box never could. - AI-driven work can produce better traceability than humans ever did — every decision and prototype connects back through the chain of thought, what Joe called "traceability on steroids." - Anyone can buy the tools — what separates enterprise work from vibe-coding chaos is policy-as-code, defined skills, and human-in-the-loop checkpoints. - Adoption happens in baby steps, not the full vision — showing people the end of the journey overwhelms them, so start at signal and layer in each next step. - The hardest part is the identity question — engineers liked writing code and designers liked crafting prototypes, and the shift is a growth-mindset problem, not a skills problem.
Links from the Podcast Endava: https://www.endava.com/ Dava.X.AI https://www.endava.com/dava-x/dava-x-ai Miro: https://miro.com/index/
Upcoming Miro Canvas Events Main Canvas 2026 landing page: https://canvas.miro.com/ Register for the streaming event on May 19 where Joe will be speaking: https://canvas.miro.com/sanfrancisco/virtual-keynote Register for the in-person event on May 19th in San Francisco where Joe is speaking live: https://canvas.miro.com/sanfrancisco Register for the in-person event on June 2nd in London where Matt will be speaking live: https://canvas.miro.com/london Register for the in-person event on June 16th in Sydney where Wesley Fagan the APAC SVP for Strategy and Chief Design Officer at Endava will be speaking live: https://canvas.miro.com/sydney
Contacting the Guests Matt Cloke https://www.linkedin.com/in/mattcloke/ Joe Dunleavy https://www.linkedin.com/in/joedunleavy/ Dave Ross https://www.linkedin.com/in/davidmross/ Dave Prior https://linktr.ee/mrsungo
Chapters
00:00 Introduction to Endava and AI Transformation 08:54 Navigating the AI Landscape: Challenges and Opportunities 17:52 The Evolution of Miro and AI Integration 24:00 Dava Flow: The AI Engagement Lifecycle 25:49 The Power of Visual Context in AI 27:03 Enhancing Contextual Understanding with AI 29:12 Temporal Dimensions in Decision Making 31:27 Navigating Change in Organizational Mindsets 35:40 Demonstrating Value to the C-Suite 38:13 Embracing the Future of Work 41:09 Amplifying Human Creativity with AI
More from Agile and Project Management - DrunkenPM Radio
Good Teams, Bad Outcomes with Martina Ziegenfuss and Jim Elvidge
In this episode, Dave Borzillo shares insights from his book 'Sanity at Scale' on how engineering leaders can shift from heroics to systemic excellence, emphasizing removing waste, visualizing work, and fostering team collaboration.
In this episode, Shalabh Tandon and Shubhada Sahasrabudhe join me to discuss their new book, 'Don't Panic Pivot: Managing Critical Issues to Prevent Crises.' We explore the differences between risks and issues, what makes an issue critical, and why they emphasize a systematic approach to managing crises. The authors introduce their nine-six framework, which includes three whys, nine ingredients, and six steps to effectively address critical issues. They highlight the role of leadership and organizational culture in fostering a proactive environment for issue management. We also dig into the intricacies of crisis management and the importance of building resilience within organizations, maintaining quality, and ensuring transparent communication during crises. You'll hear their tips on how to find balance between discipline and drama in crisis response and how AI can enhance crisis management strategies. The conversation highlights actionable insights for leaders and organizations to navigate crises effectively.
Agile 2026 is almost here! And John Tanner and I will be presenting on Monday, July 27th, at 3:15 PM. Our session is called When AI Starts Doing Your Job: Surviving the PM Identity Crisis. If you are someone who has been struggling with getting out of the gate with AI. If your hesitating because you don’t know how to start, you don’t know what to do with it, it's just too overwhelming, or you can see how it impacts the other roles, but not your own, you are the reason John and I put this session together beause that was me, 8 months ago, moaning about how I didn’t want to be the prompt guy.
I sat down with longtime friend and fellow PMI volunteer Mattias Georgson Petrén to get schooled on Half Double, the Danish hybrid framework that has been around for over ten years but I only just learned about. Mattias walks through where it came from (a frustrated group of practitioners, industrial funding, and research backing from Aarhus University), what it promises — half the time double the impact — and how it blends Agile cadence with the discipline of traditional, stage-based product management. We get into co-location, the 50% staffing rule, and why he treats Half Double as a lens layered on top of existing frameworks rather than a replacement for any of them. We close out with a conversation about where project managers fit into an AI-driven future that left me probably less worried than I've been about this in months.