I had an app that needed rebuilding, a clear idea of what it needed to do, and one familiar problem: I didn’t have several spare months to do it.
With AI’s help, I completed the rebuild in a few days.
But this story doesn’t start with AI. It starts in 2019, with a workmate, a spreadsheet, and far too much time spent chasing people for money.
A practical solution to an everyday problem
My workmate was managing our workplace Lotto syndicates. She was using spreadsheets to keep track of everything, constantly emailing members, checking payments and following people up.
It was a lot of administration for something that was supposed to be a bit of fun.
Watching her work through it, I thought: surely there’s an easier way.
So I built Lotto Syndicate Manager — lottosyndicate.nz.
The purpose was straightforward: make running a syndicate less of a chore. Bring the records together, make it easier to keep track of contributions, and keep everyone informed. Today, the app records member balances, calculates each member’s share of winnings and sends updates by email. It manages the administration, not the ticket purchases or the syndicate’s money.
Since its launch, around 100 syndicates have used the site regularly. What began as a solution for someone I worked with became a useful tool for other organisers facing the same frustrations.
That’s the kind of development I enjoy most: noticing a real problem and building something that makes life easier.
Seven years later, it was time for a fresh start
The original site was built on WordPress, with substantial customisation to turn a website into an application.
It did the job. I continued updating and maintaining it, but seven years on, the underlying code and structure were showing their age. It needed more than another round of maintenance. It needed redevelopment.
The challenge wasn’t knowing that the work needed doing. It was finding the time.
Between client projects and running Lighthouse IT, a rebuild likely to take several months was difficult to fit in. It was one of those projects that mattered, but kept getting pushed back by everything else.
Then the way I worked began to change.
From writing proposals to building applications
I started using AI in my business in 2025. Initially, my requests were fairly straightforward: help me create a presentation, write a proposal, or get a first draft onto the page.
Useful tools, but not a fundamental change to how I developed software.
Since then, my use of AI has moved well beyond writing assistance. I’m now using it to help write applications, develop automated processes and work with agents.
Compared with how I was working even six months ago, it feels like a different world.
So I decided to put it to work on a project I knew inside out: a completely new website and app for Lotto Syndicate Manager.
I supplied the specifications and watched the new system take shape.
The result genuinely surprised me. It was better than I could have produced on my own. I then spent time fine-tuning the application, refining how things worked and adding features.
A project I had expected to take months came together in days.
The new public-facing website uses Astro, a framework designed for fast, content-focused websites. The application is PHP-based and runs with my usual hosting provider.
For me, the significance wasn’t simply faster coding. It was finally being able to complete a worthwhile project that had been held back by the time it would require.
“I’ve built an app” is not quite the finish line
This experience has reinforced something else for me.
I’m seeing more people have a go at building applications with AI, and many are getting impressive results. Someone who understands a problem can now get much further towards a working solution than they might have expected.
That’s exciting. People closest to a problem often have excellent ideas about how to solve it.
But there’s a gap between “I have an app” and “other people can reliably use my app.”
Where will it run? How will it connect to a domain? What needs to happen before customers start using it? How will access, data, backups and future updates be managed?
Those are the questions that follow the first successful demonstration.
In my case, I wasn’t starting with an idea alone. I had seven years of experience with the existing application, a clear understanding of the problem and the technical background to get the replacement online.
AI accelerated the development. It didn’t remove the need for that understanding.
Where Lighthouse IT can help
At Lighthouse IT, we can help turn a practical idea into a working application, or give an existing website or app a fresh start. Custom development can mean building something new, rebuilding an ageing system, or extending what already works.
We can also help when you’ve already made a start with AI. You might have a promising prototype but need someone to assess what’s there, work through the technical requirements and provide the hosting and infrastructure needed to make it available to others.
And getting online is not the end of the relationship. Hosting, maintenance, updates, security and backups are part of keeping a service useful after launch.
Sometimes the answer won’t be a whole new app. It may be an automated process that removes repetitive data entry, follow-ups or reporting from someone’s day. That practical approach is central to our AI and automation services.
Not every project will be a few-day rebuild. But this experience has changed how I look at projects that previously seemed too time-consuming to take on.
Lotto Syndicate Manager began with a simple thought: there must be an easier way to do this.
Seven years later, that same thought helped me rebuild it.
Have an ageing website, an app you’ve started building, or a business process that takes more effort than it should? Get in touch with Lighthouse IT. Let’s turn it into something people can use.
