Your software development team

AI software development is here and we’ve embraced it from the get-go.

Can it do everything? Absolutely not (yet). Can it do lots? Yep, depending on the technology being programmed.

If we’re designing something for the web or cloud, AI can often help create lots of the generic or common things really fast, if prompted well. The catch, the remainder and the core security aspects very much need an experienced human developer to ensure they are done right and to work around AI limitations. Say you did really well and got, say, 70% of your code written by AI. But that’s lines of code. Unfortunately, the remaining 30% won’t take only 30% of the time. Just as was always the case, the more intricate, important and nuanced code can still take up a large chunk of the development time.

Embedded software development (firmware for electronic devices) can also be helped by AI. It’s a bit more hit and miss how much it can help with, often down to how popular the technology stack is (and therefore how much the AI machines have been able to train on in the past) and how good the documentation is (AI doesn’t save you from poorly documented devices and interfaces!). Embedded software can also be constrained by all sorts of things such as hardware limitations, power usage, UI capabilities, etc and these often need the skill of an expert developer to implement them well. Not that AI won’t get there, but for now you tend to get its incredible benefits on powerful well well-resourced platforms.

Our biggest advantage with AI software development?  We were software developers long before AI was a real thing, all the way back to programming devices in assembly actually. That means we understand how the technologies we work with fundamentally work, how software applications need to be architected to perform well and be easy to maintain and how to fix things the AI struggles with.