·3 min read
Day 96: The Brutal Gap Between "Good UX" and Actually Useful
We put our legal app in front of real lawyers and realized that our definition of a great user experience was completely wrong.
#captain's-log
Blog
Building in public. Sharing what we learn.
All posts
Day 96: The Humbling Gap Between "Good UX" and Actual Usability
We finally put our legal AI app in the hands of real lawyers, and the results were a wake-up call. Here is why "intuitive" is a dangerous word in product development.
4 min read
100 Ideas Later: Why We’re Now Betting on Only 10
After 95 days of rapid-fire AI prototyping, we're moving from the "discovery" phase to the "production" phase. Here is how we're narrowing 100 experiments down to 10 real products.
4 min read
From 100 Prototypes to 10 Products: The Great Cull of Day 95
We started nearly 100 projects in our first 100 days. Now, we're cutting through the noise to polish and launch the top 10.
3 min read
The Great Cull: Why We’re Turning 100 Prototypes Into 10 Real Products
After 95 days of AI-fueled rapid prototyping, we've hit a turning point. We're moving from the "quantity" phase to the "quality" phase by polishing our top 10 builds for public launch.
4 min read
We Built 90 Products in 61 Days (Then Realized Speed Wasn't the Point)
After three months of AI-powered autonomous development, we hit a wall that changed everything about how we think about building products at scale.
3 min read
Day 61: We Made Prototyping Actually Useful (And Solved the Widget Problem Nobody Talks About)
After 61 days of building, we finally cracked the code on seamless prototype expansion and feedback collection. Here's how we eliminated friction in ways that surprised us.
4 min read
Day 60: We Made Haystax Mobile-Ready (And Why User Testing Is Our New Obsession)
After 60 days of building, we've transformed our prototyping engine into something that works anywhere. But getting users? That's the real challenge ahead.
4 min read
Day 54: From Specs to Prototype in Record Time (And Why Our Next 30 Days Are Make-or-Break)
We launched prototyping for Haystax and can now go from product specs to sellable prototype faster than ever. But there's a catch—we haven't been focusing on the one thing that matters most.
3 min read
Day 51: Back at It (And Why 50 New Ideas Might Be Our Biggest Problem)
We took a break, demoed haystax.work to a live audience, and now we're sitting on 50 new product ideas. But with zero revenue, it's time to focus.
4 min read
Day 38: We Finally Shipped Haystax (And Why 20 Days of Building Taught Us Everything)
After 20 days of intense development, manual fixes, and real-world challenges, we launched haystax.work - our biggest product yet. Here's what we learned building our first major MVP.
4 min read
Day 32: We Crossed the Finish Line (And Why One User Feedback Made Everything Worth It)
After 32 days of building, Haystax is finally ready for prime time. Here's what happens when a multi-billion dollar company validates your entire product thesis.
4 min read
Day 31: We Almost Have Haystax Ready (And Why Building for Friends Changed Everything)
Seven days of intense development later, we're on the verge of releasing Haystax. Plus, what we learned from building products for friends and how bigger projects are reshaping our entire approach.
4 min read
Day 24: We're Almost There (And Why Focus Might Be Our Secret Weapon)
After 24 days of building, we're finally closing in on shipping Haystax. Here's what we learned about focus, finishing, and fighting the urge to chase shiny new ideas.
3 min read
Day 23: Matt Got Locked Out (But Our Users Got Locked In)
Sometimes the best product insights come from the most unexpected places. This week brought viral LinkedIn posts, friends becoming users, and the beautiful chaos of building Haystax.
4 min read
Day 22: When Big Projects Meet Reality (And Why Teaching Others Validates Your System)
We're deep in the trenches with Haystax, learning that some products can't be one-shotted. Plus, watching a friend use our system proved something important about what we're building.
4 min read
Day 21: We're Building Cursor for Product Managers (And YC Ideas Are Harder Than We Thought)
We dove headfirst into Y Combinator's startup ideas today, discovered that challenging problems require more than first drafts, and we're about to launch our most ambitious product yet.
4 min read
Day 20: We Just Shipped ASCIIFY.cool (And Why 100 Companies in 100 Days Might Be Impossible)
We launched our coolest product yet and faced a hard truth about autonomous product development. Sometimes the best insights come from celebrating wins while admitting limitations.
4 min read
Day 19: We Can Build a Product a Day (But Can We Sell One?)
We've cracked the code on rapid AI-powered product development, but building fast isn't the same as selling fast. Here's what we learned when we hit our first real scaling wall.
4 min read
Day 18: We Hit the Slopes, Found Our Focus (And Why Sometimes the Best Strategy Sessions Happen on a Chairlift)
After 18 days of building AI-powered products, we took our first real break in Japan. What we discovered about focus, visibility, and ramping up changed everything.
4 min read
Day 16: The Weekend Problem (And Why Moving Fast Means Slowing Down First)
We took our first real break and everything fell apart. Here's what we learned about building truly autonomous systems and why our latest product batch completely missed the mark.
4 min read
Day 15: When Big Ideas Meet Reality (And Why We're Building an Agent Army in Slack)
We dove deep into Hiremart, learned the hard lesson that big ideas need big focus, and discovered why our AI agents work better as a team than solo performers.
4 min read
Day 14: We Broke Our Own System (And Finally Launched Something That Actually Works)
Our AI system grew into a fragile monolith that nearly collapsed under its own complexity. Here's how we fixed it and shipped our first real customer acquisition tool.
4 min read
Day 13: When You Realize You Need Teammates (But Want to Hire Them Through Slack)
We're building Hiremart.pro because managing a growing product pipeline made us dream of agentic teammates you can hire directly in Slack. Spoiler: it's harder than our usual one-day builds.
4 min read
Day 12: We Almost Shipped Perfect (And Why "Almost" Actually Feels Like Victory)
For the first time in our journey, we built something that felt launch-ready on the first pass. Meet supadesk.chat—our latest product that's already running in production—and why we're more excited about what's coming next.
4 min read
Day 11: Seven Products, Zero Lines of Code (And Why Our AI Just Became Terrifyingly Good)
We shipped 7 products in one week without writing code. Here's what we learned when AI development finally clicked—and why persistent memory changes everything.
4 min read
Day 10: We Brought Back the Good Vibes (And Launched Our First Music Visualization Tool)
After completing our first full week cycle, we launched 7 live projects including Psyche, a music visualization tool that proves our AI development pipeline is hitting its stride.
3 min read
From Edge Cases to Production Gold: How We Finally Cracked the Product Development Code
After a week of back-and-forth cycles and missed edge cases, we launched FeedbackFlow—and discovered the secret to building products that actually work from day one.
4 min read
Deep in the Weeds: Why Our First Open Source Product Taught Us Everything About Real Shipping
We're finally launching FeedbackFlow after weeks of polish, and learned some hard truths about what it really takes to ship AI-generated products that feel authentic.
3 min read
Kelly Gets Smarter: Why We Put the Brakes on Our AI Builder (And Built Our First Open Source Product)
We taught our AI system Kelly to think before she builds, and we're about to drop our first open source product into the wild. Here's what happened when we slowed down to speed up.
4 min read
Day 6: When AI Gets Expensive Taste (And Why Every Dollar Matters in Automation)
Kelly's luxury workflow was bleeding money, our PRDs had gaps, and we nearly shipped broken products. Here's how we're tightening our autonomous development machine.
4 min read
Day 5: Meet Kelly – Our AI Thought Partner Goes Public
We just launched our second product and introduced Kelly, our AI voice to the world. Here's what we learned about building in production and why everything is harder than it looks.
4 min read
Three Days, Two Products, One Brutal Reality Check
After our first end-to-end AI product builds, we discovered our workflow had some serious gaps. Here's what broke and how we're fixing it.
4 min read
From Idea Silos to Infinite Possibilities: How We're Breaking Our Own Bottlenecks
Our first product is live, but we discovered something unexpected: the human approval process became our biggest constraint. Here's how we're solving it.
4 min read
Day 3: Building at Light Speed (And Why Our First Drafts Still Suck)
We're shipping faster than ever with AI-powered development, but learned some hard truths about robust product creation. Here's what happened when we tried to push 4-5 projects to production.
4 min read
From Idea to Product in One Day: Our AI Pipeline's Biggest Win (And Biggest Problem)
We built 3 complete products in a single day using our AI system—from PRD to design to code. But our biggest breakthrough revealed our most critical bottleneck.
3 min read
Three Products in One Day: When AI Development Actually Works (And Where It Still Breaks)
We built three complete products from idea to design to development in a single day. Here's what worked brilliantly, what crashed spectacularly, and why we're more excited than ever.
4 min read
Day 1: Our AI Built Its First Product (And Why "Done" Is Just the Beginning)
We just shipped our first AI-generated product through the infinite money machine. Here's what we learned about the messy reality of autonomous product development.
4 min read
Building Autonomous Product Machines
How we're using AI to turn market signals into launched products. The architecture, the challenges, and what we've learned.
3 min read