How we use AI
AI has completely transformed software engineering. But we've come to believe it's not just about software. Really, any problem can now be tackled with AI.
For years now, we've been saying that AI can, and will, go well beyond software. Just as it has revolutionised software engineering, it will help the lawyer, the researcher, the teacher, and just about everyone else with a problem to solve.
These days we work with three tools every day: Grok Bot, Codex and ChatGPT. We help a wide variety of groups, from Oxford University PhD students and financial advisors in Australia to private equity companies in the US and food manufacturing businesses. Here are two examples we're able to talk about, and why we think they matter for you, even if you've never written a line of code.
A language-learning app, from idea to three platforms
Colexi is a language-learning app. With AI, we built the web app, the Android app and the iOS app in a matter of a couple of months.
That alone would have been a big deal. But an app isn't only software. For the same app, we also wrote around 200 books, thousands of flashcards and classes, and translated the books into five languages, among other things.
None of that is code, and AI helped with all of it. That's what changed our thinking. AI isn't only a faster way to write software. It can help with the whole project, including the parts that have nothing to do with software.
A fleet of specialised bots
The way it works is simple to describe. Instead of one general-purpose assistant doing everything, we use a fleet of specialised AI bots. One looks after the Laravel side. One focuses on mobile. One handles content, and another handles courses.
A COO bot coordinates the others, so the work stays organised and moving in the same direction.
A volunteer surgical team's visit to Tonga
The second example is very different, and we think it's the more interesting one.
We helped a medical charity during a volunteer surgical team's visit to Tonga, supporting the communications around the visit. Apart from some small updates to the Tonga Hospital Projects website, this was almost entirely not a software job. And those website changes? AI did them completely.
What the work actually was
Most of it was writing and outreach:
- Five daily updates while the visit was under way.
- Two press releases.
- Finding different community organisations who would care about the project.
- Finding email addresses for news outlets across New Zealand, Australia and Tonga.
- Bringing attention to the project.
These numbers are estimates, and they relate only to the Tonga project. Before AI, we'd estimate that the five daily updates, the two press releases and the outreach would have taken roughly 20 to 40 hours of work. With AI, it came down to about 2 hours.
That's the part that still surprises us. Nothing about the task was technical. It just needed a lot of careful, tedious, useful work to be done quickly, and AI was very good at that.
It isn't just for software engineers
When people hear that we build software with AI, they often assume it's a tool for developers. We understand why, because that's where it first changed everything for us. But the Tonga project is a good reminder that the real shift is bigger than that.
You don't need to know how to code, and you don't need to understand how it works under the hood. You just need to be clear about what you're trying to achieve, and willing to try.
If a problem can be described, it can usually be tackled with AI. Writing, research, planning, outreach, organising, translating, teaching: the list keeps growing.
So here's the question we'd encourage you to ask: what problem do you have? Then try tackling it with AI. You may be surprised how much of it comes apart.
Want to talk it through?
At Camino Dev, we build custom software at a fixed price, and we use AI throughout our work. If you have a problem you're not sure how to approach, software or otherwise, we'd be glad to hear about it.