In this video, Dawid Naude explains five ways that AI can be used to make compliance reporting more efficient, and thereby reduce both the cost and time burden of reporting so that NFPs can free up their resources to focus on their purpose, and so do what they are supposed to do to make a difference.

VIDEO TRANSCRIPT

Okay. So I'm gonna go through one example of how AI can help not for profits, and this is specifically for helping them with compliance reporting.

There are five ways AI can help compliance reporting for not for profits.

The first isn't actually gathering the information required for the compliance reporting. Now this information is typically stored in systems, in emails, in spreadsheets, in people's heads, And getting all of that information out is typically or very commonly a very manual process.

But what if you could automatic message all the people that need to provide information, they can type it in over Microsoft Teams or over Slack or over an email, and we have AI that automatically interprets all of that information, and produces some kind of initial reports really automating that whole process.

The next step is when you're consolidating that and creating that initial report, and highlighting any problems that there may be. So you may have noticed a very big change between last month and this month and typically that'd be something someone would pick up manually. Whereas now, the AI will go through that data, create a report and highlight any key problems that it might think, it's found.

And when it does find a problem, what we can do as well is use AI to try and manage that exception handling process And often the exception handling process is really just speaking to a bunch of people to try to find out what happened.

So what if you could automatically message someone, SMS them, and try to find out if a particular event happened that needed to happen that you needed to report on, and they could reply to it and say, no, it didn't happen. It's been shifted to next week. And capturing that information, marking it off, or then following up again later. So actually trying to map that entire exception process end to end and seeing how much can be automated.

The next is ensuring compliance to government standards.

So there's a couple of ways AI can help here. The first is your own NFP's interpretation of the government reporting obligations.

You could put that in some kind of knowledge base that is interactive through a chatbot where you can ask it a question and say, what are my obligations on reporting for this particular thing that I'm about to do, and it would, tell you that. The second thing is you could actually have it interpret all of your activity.

So it can tell you we think that you're compliant in these nine areas, but there are these three areas that we think that you need more information to meet the minimum government standard.

And the final one is actually producing the report that's required to be given to the government for your reporting, obligations.

So being able to actually automatically create the doc document, with the right formatting, with the right headings, with all of the right information, pulling it out of all of the other sources to create the first cut of that report, you then would go through it, decide whether or not it meets it, quality check it, and then finally upload it. And to give you one idea of how we've been able to do this for another not for profit, for the information gather gathering, we had AI look at Microsoft teams, look at HubSpot and try match activity across both of them because important activity across both was captured in both of the systems.

And it was typically a manual effort to try and manage those. So we have AI basically doing that information for us. The next was it would actually create an exception list where activities that were meant to happen, weren't recorded as having happened. And we have an SMS that goes automatically to a volunteer to ask them if they'd met with, one of their clients.

They would reply and say, yes, we did, would automatically then mark that activity off as complete, or note was scheduled to next week, and then it would create a follow-up note for that.

If they hadn't replied to then somebody would follow-up with a follow-up phone call after that. The next one is then taking all of that information and creating the first version of the report for upload.

So those are five steps in the process that AI can make a really meaningful impact, on any one of them. And if you look at your own process, you'll probably be able come up with a clear example of where there's a bottleneck in your process, and I hope that this gives you some ideas of how AI might help.