Why I Built AssociationAI.ai: A Founder's Story

I’ve been experimenting with AI on and off for about a decade. I remember TensorFlow for Poets being one of the first platforms for less technical people starting out, and I loved it. I was most intrigued by its image recognition and classification tutorials, which let you work step by step in a web notebook with copy-and-paste code you could run. You’d begin with a few blocks of code and file folders, and you’d end up with a trained, working model.

The entire experience felt like a rush. I was building something I had no experience with, something that felt like it belonged in a sci-fi movie. I could work through it line by line and see the result come to life. It felt like a window into the cutting edge, giving you a framework to train and build your own. I would even swap asset and image folders, follow the provided process, and end up with what felt like a new model. These weren’t capable models, even then, but they were a great learning tool to get exposure, explore, and dream about the “what if” product ideas.

If I remember correctly, this took place circa 2017. I remember sitting at my desk, tucked away in the corner of my room in front of a window, and spending all of my free time on AI/ML coding tutorials. At that time, I was living with my mother, having moved back home after going all-in on a mobile application I had paid someone else to build. The app ultimately failed because it lacked a distribution strategy.

In hindsight, this proved to be an excellent lesson:

Success is earned, not paid for. Founders with product ideas who lack the technical prowess to bring them to fruition are themselves a “mark” to be swindled. If you cannot articulate what you’re going to build and how you’re going to accomplish your goals from a technological perspective, layer by layer, you will be taken for a ride, and it will be on your dime.

Although it’s a great lesson to reflect on now, it was painful at the time. I worked for a leading trade association publisher in my hometown of Gainesville, Florida, and had performed exceptionally well under the guidance of some excellent people that I was lucky to work with. However, I wasn’t content; something was missing. The impact I had so desperately wanted to make on the world was becoming more distant, like a fading memory of a dream: fragmented and indistinct.

By then, the world had fully transitioned to digital media, and the print membership directories I sold ads in weren’t valued by association members as they once had been, and I wasn’t in a position to improve the product; I was there to sell what they had. This wasn’t an immediate shift; it happened gradually, from 2014 onward. It ultimately led me to hire a developer to build a social media app that could serve as a digital membership directory for trade associations. I began with a consumer music app to test the framework and avoid a conflict with the non-compete agreement with my employer at the time. That brings us full circle: a failed app and living with my mom at 30.

This failure ultimately led me to begin my journey of self-education. I wanted to become more capable of making the positive impact on people’s lives that I was so desperately seeking. After three years and every dollar I had, I was left with broken, deprecated code. I needed to figure out how to rise above that experience and build something new from the ashes.

The skeleton of the app was purpose-built to hold any user data, so I figured I could take this social media platform and apply it across numerous industries, and due to my extensive work with water associations, I decided to turn it into a social network for water operators that I named “WaterAI,” which leveraged an open-source image classification model to identify equipment on-site and suggest replacements for that equipment. There was only one problem: I didn’t have access to the millions of images needed to train such a model, so I approached the water operators at GRU to see if they’d let me take photos. To my surprise, they were receptive and organized an in-person meeting with three supervisors who managed GRU’s water processing plants.

I remember arriving at the Murphy Water Treatment Plant with copies of water-industry publications I had worked on to demonstrate my experience, and with my app ready to go. I remember sitting in the conference room, surrounded by three industry experts, their eyes fixed on my phone screen during the demo as we discussed its potential and my vision for it. They all saw promise in the idea. It wasn’t fully there yet; the model mistook a chair for a coffee mug during the demo, but they could see the potential. One even said that if I got it right, he could see it becoming the de facto source of industry products, and even though it had a long way to go, I felt redeemed: a group of industry experts saw my vision and agreed that it had merit.

Then came the sobering moment: the conversation shifted to the publications I had brought with me when one of the managers said, “Hey, do you know what we do with those magazines you’ve got there when we get them in the mail?” He looked at me. Somewhat surprised by his question, I replied, “No, what do you do with them?” and his response made my heart sink into my chest: “We toss them into that recycling bin right there by the door as soon as we receive them. Anything I need, I use this.” He picked up his phone and held it up, shaking it. He reassured me that I was on the right path with the app, but that the magazines were a different story.

This was the beginning of the end of my career selling ads in the water industry. I had sunk my life into publications I thought were revered across the industry, only to hear that they were being thrown away with the junk mail. I had lost nearly all belief in the publications I was selling, except for one: the United States Geospatial Intelligence Foundation’s Trajectory Magazine, for which I led sales. It became my sole focus for the remainder of my time with my employer. Eventually, they took me off the magazine, saying other projects earned more for both the company and me. But I wasn’t in it for the money, so I quit.

In 2019, I launched Trihelix and partnered with USGIF that same year. I worked exclusively on Trajectory for four years before branching out to serve companies throughout the industry directly. It was an excellent experience, and I am forever grateful to have had the opportunity to serve such an amazing group of professionals.

This brings us to today, September 28, 2026—a good place to end this story and introduce my new project, AssociationAI.ai: a forever-free resource that allows me to give back to the community where I began my career: nonprofits and trade associations.

I built this website to help you and your teams become more proficient with artificial intelligence. Its educational articles and tutorials are researched, written, and edited by a team of AI agents under my direction, demonstrating what’s possible with the right frameworks and knowledge. I hope you find as much value in using it as I’ve found joy in building it.

Thank you,

Joseph Bradley Watkins

AssociationAI.ai