Where could AI actually help your business?

Look beyond writing

In one organisation, I developed AI applications across commercial, marketing, compliance, and information management.

These included:

• Extracting information from complex documents into a consistent format.

• Comparing transaction details with supporting records and flagging differences for review.

• Creating content from approved material, with brand standards, and channel requirements built into the instructions.

• Reviewing documents against defined criteria to identify missing information or issues needing attention.

Each started with work people were already doing: proofing, comparing, and bringing information together.

The question was where AI could save time while making sure the work was still checked properly and any decisions stayed with the right person.

Start with necessary work that’s time-consuming

Think about the tasks your team needs to do each week that are repetitive, tedious, or involve a lot of manual checking.

Do they read through documents to find the same information? Compare records manually? Spend hours gathering material before they can make a decision?

Those are places worth investigating. Before choosing a tool, ask:

• What part takes the time? Gathering information, checking it, preparing an answer or making the decision?

• What should a good result contain? Be specific about what someone needs to use.

• What still needs a person? Identify what someone must check, decide or approve.

Your answers may point to an AI application, a simpler automation or a better use of a system you already have.

Make it work for the business

For document review, I designed the application to compare information against requirements and flag issues for a person to investigate. The final compliance decision stayed with a person.

For content, it needed approved source material and clear standards. For data extraction, it needed to flag missing information rather than fill the gaps with assumptions.

The value comes from understanding the work well enough to decide what AI should do, what it needs and what a person still needs to handle.