The Gloo AI hackathon: Come Test Your Idea, Your Tools, and Yourself

The opinions and views expressed in Perspectives content are the author’s own and do not reflect the views, values, or position of Gloo.

____

The Crossroads

A year ago, I was between jobs, standing at the fork every builder eventually reaches: go back into the comfortable big tech corporate world, or bet on building a new startup. When I reflect on my work in big tech, the primary goal of the work was getting people to click on more ads. For the startup path, what would the north star be? When I came across the Gloo AI Hackathon, their framing of building technology for human flourishing instantly resonated with me and showed me what my next chapter should look like.

Hackathons have always been home turf for me. In a prior life, I was the all-time hackathon champion at Yahoo with 13 wins. So when I needed to answer a life-sized question, going back to a hackathon felt like the natural place to do it.

If I was going to do the startup, the next set of questions came up.

• How many people would I need to get this thing off the ground?

• If we leverage AI for coding, how capable are the coding agents?

• If we leverage AI to solve problems, what can it actually do, and how well?

Why I Took It to a Hackathon

Instead of agonizing over the decision in a notebook, I decided to answer it the way I answer most hard questions: by building something ambitious and seeing what happened. The hackathon resolved the big questions quickly because I time-boxed the experiment to one week, and at the end of the week I had to demo something to judges and an audience. Action beats waiting for the perfect time.

In terms of human flourishing, I knew there were two dimensions I wanted to tackle.

1. Finances: financial stability and freedom from money-related stress

2. Purpose: sense of meaning, calling, and direction, often tied to work

I had this grand vision that AI could help reduce the cost of goods and services in the real economy, and make those services more abundantly available. Think housing. What if we could move approvals through faster? What if we reduced the cost of housing by cutting out mistakes and busy paperwork? If we could reduce the drudgery of that paperwork, could we also improve people's sense of purpose at work? Think financial services. What if you could get financial help at a tenth of the cost?

The Gloo AI Hackathon forced me to think about the minimum viable demo that could show this grand vision, and that starts by finding a problem. In my group of friends, we were talking about a goal of reducing our credit card debt. My friend Antonio also happened to be visiting, and he used to run a credit repair business, so I took the opportunity to interview him about how his business ran and where the bottlenecks were. I ended up with his playbook on how credit repair is done.

Test Your Idea

Before I wrote a single line of code, I wanted to pressure test the idea. I wanted to validate the credit repair playbook. I wanted to talk to people who needed help with their credit, and find out whether the solutions felt manageable and understandable to them. I talked with financial coaches to see if they could use a tool like this. I also wanted to assess the tools that people were currently using, and what was working and what wasn't.

One key learning from the financial coaches was that people were more likely to succeed in their financial goals if they had a financial accountability buddy who met with them monthly. Nobody wants to get on a call and say they didn't hit their goals for the month.

Test Your Tools

The second thing I wanted to know was how far AI had actually come. I had to separate the hype from the reality. I wanted to test how good coding agents were across different domains such as design, frontend work, backend development, architectural design, and testing. A hackathon is an ideal place to run that test, because you start greenfield, unburdened by legacy code. Nothing muddies the read on how good the tools really are.

Then the next test was the agent's capabilities. Could it accurately read financial information off a bank statement? Could it draft a PDF letter and follow design guidelines? Could it provide a sound financial roadmap given the user's unique circumstances and context?

In past hackathons, I would typically compete with a team of four. For this one, I entered as a team of one to see how much AI could amplify my efforts.

Test Yourself

The third test was the most personal one: I had not written code in about five years, and I genuinely did not know if I still had the skills.

To my delight, I found myself working with coding agents much like I used to work with engineers as a principal engineer. In that role, the main responsibility was to scope out work, delegate it to junior engineers, and review their changes and provide feedback. That workflow is exactly the same with coding agents, and I found myself very much at home. It let me focus on the big picture of what I was trying to build instead of wrestling with code syntax.

I was also impressed that coding agents could act as a personalized tutor. I'm slightly ashamed to admit I had never built a continuous integration pipeline in my life, because I always had the luxury of build engineers in my past jobs. This time I had to build one myself, and the coding agent walked me through the key concepts. We had one up and running fast.

That experience convinced me of something about what it means to be a builder in 2026. The roles of engineer, designer, and product manager are collapsing into one. You have to wear all three hats at once. When everyone can generate code, the constraint is no longer output; it is judgment. You have probably heard the term AI slop. You can build endless things, but if they are the wrong things, or poorly designed, they never find traction.

What Five Days Actually Produced

What came out the other end was Kommonwealth, an AI financial coach: you feed it your bank statements and credit reports, and it hands you a prioritized to-do list for improving your credit score.

The key differentiator was that the coach didn't just give you suggestions; it could take real-world actions on your behalf. Most people at that point were using AI like a search engine to find answers. For this financial coach to be amazing, it had to be an agent that could reach out and touch the world. 

Things that Kommonwealth could do: 

• Draft a credit dispute letter and actually mail it on your behalf with one button click. 

• Schedule a monthly call with your financial accountability buddy to help keep you on track. 

The design inspiration came from reading Atomic Habits, where the idea is to reduce as much friction as you can for positive behaviors.

The Mission Ahead

I won the developer track, and it came with a healthy cash prize that I turned around and used to help bootstrap my startup, Kommon.

The Gloo AI Hackathon was a weekend. The mission is not. It turns out that if you can build an agent to handle personal finances, you can build one to handle a company's finances and operations too. That is the company we are building now: the AI back office for the firms that keep the physical world running, one company at a time. This means construction, home builders, field services, logistics, and manufacturing. I care about these sectors because they underpin our world. We have a crisis in affordable housing and affordable services. I believe AI can make those things cheaper by driving down the cost of delivery.

At my last company, I was forward deployed to construction job sites and factory floors, and almost everything still ran on pen, paper, and spreadsheets. Construction is the second least digitized sector in the entire economy, behind only agriculture. That gap is exactly where AI can give people their time back. Our customers, usually a CEO or CFO, live through an end-of-week crunch where the books have to be reconciled by working over nights and weekends. We automate that work and compress it from hours to minutes. The tagline we keep coming back to is simple: we help people get their nights back. 

Find Your People

The prize money was real, but the thing I would come back for is the people in the room, because the talent density at a Gloo event is unusually high. One of Gloo's superpowers is convening world-class speakers, industry leaders, subject matter experts, and passionate builders. There's incredible value in the conversations you have and relationships you build there.

If anything, those human connections matter more in an AI-native world, not less. I learned this the hard way. At a conference recently, someone wanted to meet about go-to-market strategy, so we set it up. I sat down, said "Hi, I'm Chris," and the reply was "Who are you?" I realized that the whole conversation leading up to the meeting was with a chatbot, not with them, and that pretty much skunked the deal. When everyone can build, and everyone can automate the first touch, the person who actually shows up and builds a real relationship is the one who stands out. A hackathon is a room full of people who showed up.

I’ve been showing up to these types of events for eleven years. I met some folks from Gloo at the Code For The Kingdom Hackathon in Silicon Valley, and since then I’ve met so many amazing people through this community, including mentors, investors, customers, and lifelong friends. One of them even officiated my wedding. Come find your people. 

Just Do It

If you have never been to a hackathon, my advice is simple: just do it, and let your version one be deeply embarrassing. The worst outcome is to build quietly for a year, launch to the world, and only then find out nobody wanted it. The moment you ship something, even something rough, you start getting the feedback that makes the thing good.

So come as a learner, a mentor, or a builder, and you will leave with more than you brought. Come test your idea, your tools, and yourself. Happy hacking!

____

This is a guest post written by an independent founder and former Gloo AI Hackathon winner. The tools and concepts discussed reflect personal hackathon projects and do not constitute financial, credit, or legal advice. Gloo does not sponsor, endorse, or verify the third-party platforms mentioned.