Turning skincare chat into real routines

We just introduced routines in Klynn.

But we didn’t treat it as a typical feature. We built it as a system that works together with AI.

Klynn doesn’t rely on a fixed product database. It runs on a real-time analysis engine. That makes building a meaningful routine system a very different problem.

Most routine apps are static. But in reality, routines evolve.

So instead of building something static, we built something that adapts, with chat at its core.

What this means in practice:

• Create routines directly in chat

• Update and refine them through conversation

• Ask AI to analyze your existing routines

• Track what you actually complete, day by day

The key difference is context.

Klynn already uses your profile, product history, and behavior. On top of that, the memories system allows it to build on what it already knows about you. So routines aren’t generic, they’re shaped by your actual usage and preferences.

We also shipped Chat v3 alongside this. It’s a more capable system overall, and the routine layer is built on top of it.

This ended up being one of the hardest things we’ve built. Especially doing all of this without relying on a static database.

Routines are no longer just a feature in Klynn. They’re part of a system where chat, memory, and real-time analysis work together.

If you want to try it:
https://klynn.app