Xami AI
July, 2026
The more CRMs I explored, the more I found myself asking the same questions.
Where did I save that note?
Which page was that deal on?
Why do I have to leave this screen just to reply to an email?
At first, those moments didn't seem important.
The more I noticed them, the harder they became to ignore.
I started wondering if the problem wasn't the features.
Maybe it was the navigation.
Xami AI became my attempt to explore that idea.
This isn't a blueprint for the future of CRM. It's simply an exploration of what happens when software spends less time telling you where to go and more time helping you do what you came to do.
The first screen.
I kept redesigning the home screen.
No matter what I added, it still felt like the wrong place to begin.
The more I thought about it, the more I realized people don't open a CRM because they want to look at a dashboard. They open it because they have something they want to do.
That was the moment I stopped designing a home screen and started designing a conversation.
Waiting.
I never liked how AI assistants just sit there waiting for you.
The hardest part of any conversation is knowing how to begin. That made me think the assistant shouldn't just wait for a prompt. It should help people find one.
Changing your mind.
I found myself switching between Kanban and lists all the time.
Not because one was better than the other, but because they helped me think about the same work differently.
It didn't feel right for the interface to make that switch feel like starting over.
Losing momentum.
I didn't think much of it at first.
I'd open a deal, check a few details, leave a comment, then spend a few seconds finding my way back to where I was.
After a while, I realized I wasn't thinking about the comments anymore. I was thinking about everything that happened before and after them.
Another tab.
I kept noticing how often I had to leave the CRM just to answer an email.
It always felt strange. The conversation was part of the work, yet it lived somewhere else.
I wanted replying to someone to feel no different than updating a deal or leaving a comment.
Getting started.
I realized writing the email was rarely the difficult part.
Starting it was.
That changed how I thought about the composer. Instead of feeling like another form to complete, I wanted it to disappear behind the conversation itself.
Picking up where I left off.
I never liked how replying to an email felt like switching tasks.
The conversation was already there. The context was already there. It didn't make sense to leave all of that behind just to write a response.
The things in between.
Not every idea is ready to become a task.
Some start as half-finished thoughts, meeting notes, or something you want to come back to later.
I wanted those moments to feel just as important as everything else inside the CRM.
Getting out of the way.
The more I worked on the editor, the less I wanted it to feel like an editor.
Formatting, links, colors, and everything else should be there the moment you need them, then disappear into the background so the words remain the focus.
The first step.
Creating a task should feel like the easiest part of the work.
I kept stripping things away until only the essentials were left. Everything else could wait until it was actually needed.
That simple change made creating tasks feel much less like administration.
Looking a little closer.
I used to think the board was where the work happened.
After spending more time with it, I realized the board was just an overview. The real work lived inside each task, where conversations unfolded, decisions were made, and progress slowly took shape.
Where it belongs.
For a while, I treated the assistant like every other feature.
It had its own page, its own place in the product.
Eventually, that stopped making sense. If AI is there to help, it shouldn't ask you to leave your work just to find it.
What I came away with.
I started Xami AI thinking I was redesigning a CRM.
Somewhere along the way, it became something else.
I found myself spending less time thinking about dashboards, navigation, and modules, and more time thinking about momentum. Every interruption, no matter how small, pulled people away from the work they had already started.
That changed almost every decision I made.
The conversation became the starting point. Notes became part of the work instead of somewhere ideas went to be forgotten. Email stopped feeling like another application. The assistant stopped being another destination.
I don't think people enjoy using software because it has more features.
I think they enjoy it because the software quietly lets them focus on what they came to do.
Whether Xami AI is the right answer, I honestly don't know.
But exploring that question changed the way I think about designing products.