This tool built with Lovable might reach 1.6M users
Jun 11, 2025
I started playing with AI and no-code out of curiosity. A few months later, I was building tools my team actually used. These are the lessons and bumps along the way.
A while ago, I started experimenting with AI and no-code tools during weekends. I had no big plan, just a lot of curiosity. I tried a few tools: v0, Lovable, and Bolt.
What worked and what didn’t
v0 was great for UI, but I struggled to connect it to real data or logic. I could design, but I couldn’t launch.
Bolt could spin up apps fast, but it hallucinated features and broke layouts. I didn’t feel confident putting anything live.
Lovable turned out to be a good balance. I could move quickly, test ideas, and ship something that looked and felt like a real app.
My first project: Manual of Me
This started during a team offsite. All the designers were together, and we were casually sharing how we like to work. I thought: what if there was a simple tool to capture all of this?
So I built Manual of Me, a quick tool where each person could share their working style and preferences. It became a fun way for us to understand each other better.
I kept building
Next, I made a CPR rhythm game. It let you tap to the beat of a song that matches the CPR rhythm. The idea was to train auditory and muscle memory for emergencies. Simple but useful.
Then I started thinking about real problems at work.
Spotting a real pain: Inventory
I noticed that many teams internally and the users externally were struggling to manage physical inventory. They didn’t know:
Who had what
Where it was
If it needed maintenance
Everything lived in spreadsheets or long email threads.
So I built an Inventory Management App using Lovable.
It had:
Map view
Address validation
Real-time photos
Check-in and check-out tracking
I shared a demo, and three teams asked to pilot it.
Talking to developers
This is when I realized it wasn’t just a weekend project anymore.
I showed it to a few developers. One said they wouldn’t even review it because it was AI-generated.
Another dev looked at it and said:
“This is solid. If I had to build it, I’d start from this.”
That gave me the confidence to keep going.
From prototype to something real
I interviewed the teams who wanted to use it. I mapped out what they actually needed.
I realized some of the features were nice-to-have. But there was one core use case everyone cared about.
So I focused on that.
Now I’m working closely with one team to personalize the app for them.
They’ll use it during a customer visit this summer. Based on that feedback, we’ll iterate.
If it works, this could scale to 1.9 million users across the org.
Helping others start
I noticed that designers around me were still hesitant to try AI tools.
So I ran a short hands-on session: Let’s build something together. During Design Summit in Cabin.
I shared my journey and asked them built something with me.
I also shared my experience in the AI Sparks program with people.
My goal was to lower the entry barrier, to make AI feel useful.
What I’ve learned
AI and no-code won’t replace developers. But they help designers and builders like me:
Validate ideas faster
Share working demos earlier
Get real feedback before investing too much
I’m still careful. These tools have limitations—security, stability, long-term support.
But they’re perfect for the exploration phase—when the biggest risk isn’t bad code, it’s building something no one wants.
That’s what no-code + AI helped me avoid.
A while ago, I started experimenting with AI and no-code tools during weekends. I had no big plan, just a lot of curiosity. I tried a few tools: v0, Lovable, and Bolt.
What worked and what didn’t
v0 was great for UI, but I struggled to connect it to real data or logic. I could design, but I couldn’t launch.
Bolt could spin up apps fast, but it hallucinated features and broke layouts. I didn’t feel confident putting anything live.
Lovable turned out to be a good balance. I could move quickly, test ideas, and ship something that looked and felt like a real app.
My first project: Manual of Me
This started during a team offsite. All the designers were together, and we were casually sharing how we like to work. I thought: what if there was a simple tool to capture all of this?
So I built Manual of Me, a quick tool where each person could share their working style and preferences. It became a fun way for us to understand each other better.
I kept building
Next, I made a CPR rhythm game. It let you tap to the beat of a song that matches the CPR rhythm. The idea was to train auditory and muscle memory for emergencies. Simple but useful.
Then I started thinking about real problems at work.
Spotting a real pain: Inventory
I noticed that many teams internally and the users externally were struggling to manage physical inventory. They didn’t know:
Who had what
Where it was
If it needed maintenance
Everything lived in spreadsheets or long email threads.
So I built an Inventory Management App using Lovable.
It had:
Map view
Address validation
Real-time photos
Check-in and check-out tracking
I shared a demo, and three teams asked to pilot it.
Talking to developers
This is when I realized it wasn’t just a weekend project anymore.
I showed it to a few developers. One said they wouldn’t even review it because it was AI-generated.
Another dev looked at it and said:
“This is solid. If I had to build it, I’d start from this.”
That gave me the confidence to keep going.
From prototype to something real
I interviewed the teams who wanted to use it. I mapped out what they actually needed.
I realized some of the features were nice-to-have. But there was one core use case everyone cared about.
So I focused on that.
Now I’m working closely with one team to personalize the app for them.
They’ll use it during a customer visit this summer. Based on that feedback, we’ll iterate.
If it works, this could scale to 1.9 million users across the org.
Helping others start
I noticed that designers around me were still hesitant to try AI tools.
So I ran a short hands-on session: Let’s build something together. During Design Summit in Cabin.
I shared my journey and asked them built something with me.
I also shared my experience in the AI Sparks program with people.
My goal was to lower the entry barrier, to make AI feel useful.
What I’ve learned
AI and no-code won’t replace developers. But they help designers and builders like me:
Validate ideas faster
Share working demos earlier
Get real feedback before investing too much
I’m still careful. These tools have limitations—security, stability, long-term support.
But they’re perfect for the exploration phase—when the biggest risk isn’t bad code, it’s building something no one wants.
That’s what no-code + AI helped me avoid.