Nicole Baer

things that tend to have a right way of doing them.

And those tasks typically have been very human intensive because it required cross-checking things like limited partner agreements to make sure the way that an expense was being allocated was in line with what had been promised to limited partners, essentially the way that a fund needs to be run.

And so we've used AI to really be able to combine all of the human knowledge of context that was through years of doing this work with people and be able to build that rich context into the agent so that it can just do those things automatically.

And it's not just saving us time for the people that were doing all of that work on sort of the back end of AI.

of our fund admin platform, but it also saves time for our customers.

And that's where we really focus a lot of our effort.

We have configuration agents that allow for

CFOs to change the way that we calculate certain metrics in the platform.

Like they want metrics calculated a certain way and they can actually set that calculation.

It sort of stays that way until they change it.

Yeah, we have agents that run in the background that essentially are always looking for

you know, I kind of call it like the janitor who works at night, like, you know, cleaning up the school and everything looks great the next day.

It's sort of the same idea.

These agents run in the background and find errors, things like, well, your balance sheet says this, but your income statement says this, you know, and we'll flag those things that sometimes are the result of, you know, human error when you move out of a spreadsheet, which is what a lot of our customers do, move out of like a spreadsheet for how they're administering their funds and move it

into software, the benefit of doing that is actually accuracy.

And that's the reason that so many of our customers move on to our fund admin platform and out of spreadsheets.

And then we also have an insights agent that allows for customers to query our data warehouse using natural language, which is really cool.

So those are the four things that are sort of in production.

And really, they're just, I mean, they're incredible tools for our customers.

And there's lots more to come.