Got AI, and don’t know what to do with it?

We are having a lot of conversations with our clients, business owners, consultants and contractors recently about what to use AI for. Some have been given access to Copilot, others have built in-house AI trained on their own datasets (props to these guys!). But a vast majority don’t know where to start.

✅ Start with a use-case! ✅

What problem do you want to solve, what process area do you want to streamline, what maze of un-structed data do you wish to understand? You don’t buy a tool and then look around for somewhere to use it. AI should be no different (though we do encourage experimentation).

If you have an idea of what you want to do, then seek to understand what data you have available. You may need data to train your model, or you may deploy a trained model to interpret your data. Either way, you really need to understand your data.

We (in construction) have historically been terrible at collating full and accurate project records, project information, project DATA! You know the scenario, it’s the end of the job, the money is running out, people are leaving or being deployed onto the new, shiny projects, nobody has been tasked with tidying up or is capable of doing so.

Well that has to change and, thankfully, is changing.

Ideally you will set up your information management systems at project inception and will be able to compile a clean, tagged and useful dataset as your project progresses. Alternatively, you can work to improve the quality of your existing datasets, augmenting where there are gaps, and preparing it for the chosen use.

Once you have these (use case and data), then you can select the right AI tool.

❓ We are currently running a few projects around risk and AI, but we would really like to know what your current use-cases are. ❓

Comment below, or message us (for anonymity if required), or email for a no-obligation conversation.

Help us to improve the industry’s capability in the use of AI and DATA.

Next
Next

Everyone agrees we need to do better. But where do we start?