HUMAN-LED · BUILT WITH AI

Ideas from a notepad.
Now working software.

There is still a human touch behind all of this. LinesFly and HuddlePops started with problems I understood, choices I made, and ideas grounded in my own lived experience. AI helped me turn those ideas into working software.

01 THE HUMAN PART

I started with problems
I understood.

I had a few ideas, but I focused on problems that would still exist even as AI keeps getting better. I wanted the products to be grounded in reality and in experiences I actually had.

I tested a few smaller things to see what AI could do. Then I turned that practice toward real problems I wanted to solve. I’m not an expert. I’m still figuring out how a lot of this works. These two products came out of those experiments.

01 LIVE + WORKING

LinesFly.com

A live queue for walk-in businesses. Customers scan a QR code, join the line, see where they stand, and get called when it’s their turn. Staff get one clear view of who’s next.

LinesFly is not the only queue management product in the market. There are plenty of other systems. This is my way of solving the problem: simple for the customer, clear for the staff, and built around the phones people already carry.

THE PROBLEMA clipboard, repeated questions, crowded waiting areas, and a lot of guessing about who is next.

MY VERSIONA real-time waitlist that lets people join by QR code, see their place, and wait anywhere.

ROUGH ESTIMATE · CUSTOM BUILD WITHOUT AI$180,000–$350,000

I asked ChatGPT what it could cost to recreate the current scope with a traditional product team. That estimate accounts for:

  • Product and UX design
  • Front-end application development
  • Back-end, database, and real-time infrastructure
  • Testing, security, deployment, and quality assurance
  • Rounds of revisions, fixes, and feature changes
This is an AI-generated comparison based on the current feature set, not a contractor quote.
Explore LinesFly ↗
BUILT WITH AI
SHIPPED BY HECTOR
linesfly.com
LinesFly, an AI-built real-time waitlist SaaS, displayed in a tablet frame
02 LIVE + WORKING

HuddlePops.com

A simple way to make meetings interactive. Add a poll, word bank, feedback question, or icebreaker; share one QR code; and let everyone answer from their phone.

HuddlePops is not the only polling or meeting engagement product. There are many others. This is the one I wanted to build, shaped by years of leading teams, facilitating groups, and trying to get more voices into the room.

THE PROBLEMAsk the room a question and hear from the same two people, or nobody.

MY VERSIONOne interactive moment that lets everyone respond from a phone and gets more voices into the conversation.

ROUGH ESTIMATE · CUSTOM BUILD WITHOUT AI$200,000–$500,000

I asked ChatGPT what it could cost to recreate the current scope with a traditional product team. That estimate accounts for:

  • Product strategy and interaction design
  • Front-end application development
  • Back-end, database, and live response infrastructure
  • Testing across devices, security, and deployment
  • Rounds of revisions, fixes, and feature changes
This is an AI-generated comparison based on the current feature set, not a contractor quote.
Explore HuddlePops ↗
NO APP
NO DECK REBUILD
huddlepops.com
HuddlePops, an AI-built interactive meeting SaaS, displayed in a tablet frame

Human-led. Built with AI.

You can click through and try them. These are working products, not just screenshots or ideas sitting in a notepad. I built them.
02 THE WORK IS STILL HAPPENING

These are the beginning, not the finished story.

LinesFly and HuddlePops work, but they’re still prototypes. I’m iterating, finding bugs, and figuring out what happens when something breaks.

I want to share that part too, not just the cleaned-up version at the end.
01 · BUILD

Keep making the thing better.

Work through the bugs, question the choices, and keep closing the gap between the idea and the product.

02 · SHARE

Document what actually happens.

The mistakes, costs, fixes, tools, and small wins while they’re happening.

03 · SEE

Eventually, put it in front of people.

Promote it, listen, and find out whether anyone wants it. If nobody does, that’s useful information too.

03 THE NEXT PART OF THE EXPERIMENT

Can one person build it
and keep it running?

Getting working software online is one thing. Maintaining it once real people use it is another.

LinesFly and HuddlePops will need attention, judgment, and care. AI can help me do more, but it does not remove the human responsibility behind the product.

01 · PRODUCT CARE

Fix what breaks.

Watch for bugs, keep the software reliable, protect the data, and make sure new changes do not create new problems.

02 · REAL PEOPLE

Listen and support.

Answer questions, understand where people get stuck, and decide which feedback should shape the product next.

03 · OPERATIONS

Run the business.

Manage hosting costs, billing, privacy, documentation, and all the unglamorous work around the software.

04 · GROWTH

Find out if people want it.

Explain the value, reach the right people, learn what earns trust, and see whether either product can become a sustainable small business.

THE QUESTION UNDERNEATH ALL OF THISCan AI help one person do work that once required an entire small team?

Maybe it makes $100. Maybe $1,000. Maybe it flops. I do not know yet. Building it was the first experiment. Keeping it useful is the next one.

Want the thinking behind the builds?SEE THE VERY UNOFFICIAL PROCESS ↗