Project Note
AI in Practice
The public version of confidential enterprise work: helping AI move from interesting idea to useful workflow, with the data, controls, delivery habits, and adoption path connected.
Problem
AI interest was not the hard part. The harder work was figuring out where it genuinely fit, what data it needed, how it should be governed, and how to make it useful inside work that already had owners, deadlines, habits, and risk.
Role
Worked across senior stakeholders, business teams, and delivery teams to sort useful ideas from distracting ones, connect them to data readiness, and create enough structure for teams to build without turning the work into theater.
What can be shared
- 90%+ AI use across enterprise functions.
- 40%+ of the organization actively building with AI.
- 12+ data sources integrated across cloud and on-prem systems.
- 1B+ rows unified into governed, accessible data foundations.
What shipped
- A phased AI delivery approach tied to real workflows.
- Prioritized use cases with clearer owners, data needs, and success conditions.
- Data readiness materials that helped teams understand what the work required.
- Useful internal tools and patterns that could be reused instead of rebuilt from scratch.
What mattered
- Making the work legible enough for teams to use.
- Keeping governance and delivery connected instead of separate tracks.
- Creating a repeatable way to move from idea to working tool.
- Staying focused on practical usefulness, not demos for their own sake.
Working themes
- AI workflow design: turning broad interest into specific work people can actually do.
- Data readiness: getting the foundations, ownership, and controls into shape.
- Adoption: helping people build confidence through tools they can use in their own work.