As AI becomes part of content, marketing and development workflows, which qualities of Xperience by Kentico become more valuable in your experience, and why?
For me, one of the biggest changes is that AI makes the quality of the underlying structure much more important.
A good content model gives people, applications and AI agents a shared understanding of the content. Clear content types, meaningful relationships and consistent metadata make it easier to interpret what content means, how it can be reused and which actions are appropriate.
I see the same with code. AI can produce working code surprisingly quickly, but working is not the same as fitting the intended architecture. A codebase with consistent patterns, clear boundaries, good naming and reusable examples gives an agent a much better chance of creating something that developers would also consider the right solution.
This is where APIs, documentation and MCP servers become important. They can provide AI with product-specific context and approved ways to perform actions. Without those guardrails, an agent may invent its own solution. It might, for example, write directly to the database because that technically works, while bypassing APIs, validation, security, upgradeability and the conventions of the platform.
In my experience, this is where KentiCopilot becomes especially relevant. Rather than leaving AI to find its own way through an implementation, it can provide the product-specific context, skills, MCP servers and supported patterns that guide it throughout the development process.
That guidance is also important for developer trust and adoption. Developers still need to understand and take responsibility for the code that AI produces. If generated code looks unfamiliar, bypasses established patterns or is difficult to explain, there is a good chance that it will be rewritten, even when it technically works. At that point, much of the initial time saving is lost.
By helping AI stay within established Xperience by Kentico development patterns, KentiCopilot can make the output easier to review, explain and accept. Structured content, predictable APIs, clear development patterns and governance remain the foundation. KentiCopilot helps reinforce those guardrails, reducing the risk of technically working but architecturally inappropriate shortcuts.
This makes the outcome easier to maintain and evolve, while also building the confidence needed for broader adoption of AI-assisted development.
I am interested in how others see this.
Which qualities of Xperience by Kentico become more valuable as you introduce AI into your workflows? And what helps you ensure that AI follows the intended architecture instead of simply finding a way that works?