The problem
What was broken before AI
Many companies tell employees to use AI, but leave the actual behavior change vague. People may have ideas for automations, but they do not know who can build them, whether they are safe to ship, or how to get them prioritized. Non-engineering teams can be especially blocked: they may see the opportunity clearly, but lack the templates, permissions, or confidence to turn the idea into something usable.
What changed
What the use case made possible
Delight.ai treated internal AI adoption like a product. Employees can create quests for automations they need, builders can form teams around those requests, and completed work earns XP and rewards. Secure app templates give non-engineering teams a safer path to production, while usage dashboards show where people are on the journey from beginner to advanced AI user. John also models the behavior personally with tools like The Gardener, an agent that organizes and enriches his notes.
Why this matters
Why this use case is worth studying
John’s approach is useful because it handles the culture layer of AI adoption. The hard part is not only getting access to models; it is creating habits, incentives, trust, and safe paths for people to build. The marketplace gives ideas a place to go. The templates reduce fear. The dashboard makes progress visible. The rewards make participation feel social instead of bureaucratic.
Use this when
When this pattern applies
Use this pattern when people across a company have AI ideas, but no clear path to turn those ideas into safe, useful internal tools. It works especially well when adoption has moved beyond curiosity and the organization needs systems, incentives, templates, and visible examples to make AI part of everyday work.
