First 5 Steps to Leverage AI in Your Business

First 5 Steps to Leverage AI in Your Business

Since the IO and Gemini announcements it’s clear this hype train has no breaks. Before delivering the previous tranche of promises, the tech companies are full steam ahead into the next wave of capabilities.

If you’re a business leader with FOMO and finding it hard to separate hype from reality here are the first 5 steps you can take to start leveraging AI with minimal risk:

1.        Educate yourself on the state-of-play - using unbiased sources. Look for authors and publishers who do not have financial incentives to sell you AI products. Stanford University Prof. Andrew Ng, Princeton University Prof. Arvind Narayanan, Gartner, SemiAnalysis and METR.

2.        Educate your staff on the theory and practice of using common off-the-shelf tools like Gemini and Claude.

3.        Create a list of 3-5 diverse use cases ranked by their value to your business. Ensure their success can be measured by tangible business KPIs.

4.        Identify which of these use cases have a large amount of data (>1M data points) already recorded and whether that data is helpful. A good way to identify this is to ask the question “if an expert in this field had this data, and infinite time, could they perform the task effectively?”

5.        Build a well-rounded team comprising a mixture of AI scientists/engineers, subject matter experts, project management and IT support staff to build the PoC.

From my experience, steps 4 and 5 are where businesses get stuck. Often, the data is in a bad state due to manual entry, platform inconsistencies or simply a lack of data recording processes. In which case, an initiative to start recording events/workflows, in a standardised manner, must be established. This will shift the timelines out, but it’s better late than never!