IGHS65 - Ryan Gerhardy, CEO at Pitchly

Jim Merrifield (00:00.763)
Well, hello and welcome to the InfoGov Hot Sea. I'm your host Jim Merrifield and with me today is Ryan Gerhardy, CEO at Pitchly. Welcome, Ryan.

Ryan Gerhardy (00:10.03)
Thanks Jim, thanks for having me.

Jim Merrifield (00:11.449)
Yeah, it's great to have you in the hot seat. to know you a little better, you and your company and what you're doing these days at Pitchly around AI and in the legal market. So why don't we kick this off by having you provide a brief introduction of yourself, your current role and one fun fact about yourself.

Ryan Gerhardy (00:30.478)
Yeah, I would love to. So I started my career in Australia, actually in equities research and investment banking, and then moved to the US in 2014, or I like to say 11 winters ago, for kind of obvious reason, and worked at a VC fund here in Des Moines, Iowa, before co-founding Pitchly So my current role, CEO of Pitchly, we have 18 team members throughout North America.

Customers use Pitchly to design agents to create, update, and publish content for the professional services industry. So we'll dive into that today, currently customers have 551 agents managing 7.8 million pieces of content accessible across eight endpoints and with live updates from 6.1 billion data points. So starting to compound and see that big data piece, which is great.

Jim Merrifield (01:24.379)
That is a lot of big data, lot of numbers there to unpack.

Ryan Gerhardy (01:27.758)
A lot of numbers to digest, and I'll go into that a bit. But one fun fact is I play ice hockey, which I feel is very strange for an Australian living in Des Moines, Iowa, but I love it. It keeps me kind of aggressive off the work site and a little bit more fun with friends on the ice. And it feels very relevant after the Four Nations tournament this week too, which was really cool.

Jim Merrifield (01:50.725)
Yeah, for sure. Who's your team?

Ryan Gerhardy (01:53.816)
I mean, I'm between the Detroit Red Wings, because they're the only team that I could watch in Australia, and the New York Rangers, because I'm always in New York for work. So I get to see them more often than anyone else in person. So I'm torn. I'm torn.

Jim Merrifield (02:05.188)
All right. Awesome. Go Rangers, go Rangers. Sweet. So let's talk a little bit about your company. What inspired you to start your company and how has the vision of the company evolved over time?

Ryan Gerhardy (02:21.614)
Yeah, it originally started really to scratch my own itch. Working as an investment banker saw the repetitive design and the time it took our team to do that was really troublesome, right? It kept us away from the analytical work, the thing that frankly made us money. And instead we were fulfilling marketing requests or pitch book submissions or award submissions. I think there's a lot of people that will resonate in your audience about that.

And really, if you say that another way, it's graphic designers, amateur designers, even attorneys or bankers or practitioners, eliminating the time they spend refining pixels, aligning, reformatting, ordering objects in slides and documents. It should be automated. And it is with Pitchly today, which is great.

Jim Merrifield (03:10.298)
For sure, I love the name Pitchly, having a marketing background. you're right, in the law firm world, it's a heavy load on those business development and marketing individuals. I mean, if you got 25 practice areas and about 300 plus or maybe even thousand lawyers, you're essentially supporting a thousand separate companies, or at least 25.

Ryan Gerhardy (03:35.33)
Yeah. And it's small teams, right? And so I think as our, you asked how the vision has evolved, right? Really we've moved into doing more work to enable these small teams and being considered kind of that BD resource. And now with the way we talk about Pitchly as an agent that you're deploying, it's very much being part of their team.

And you deploy the agent to complete the private equity, banker or lawyer or attorney bio, right? And so you're deploying these agents at a use case, at a content level, and then at a document level, so they're done autonomously.

Jim Merrifield (04:17.968)
That's awesome. definitely going to, I think at the end of this conversation, I'm probably going to ask for my demo just because I'm intrigued there for sure. So what do you see are the biggest challenges or opportunities for AI and information governance in the tech industry over the next, say, I even want to say five years, maybe the next several years?

Ryan Gerhardy (04:24.386)
Yeah. Yeah, for sure.

Ryan Gerhardy (04:42.958)
Yeah, and I have a few examples here. So we've launched our first product back in 2018, right? And it was a very simple, you fill out a form, you get content. And kind of where we've come and where we've evolved is there's more configurations, integrations, automations. And the way we look at kind of where AI is going is we look at the three kind of layers of data management. So...

Wendy Batchelder, one of our board members, the former chief data officer of Salesforce, kind of educated Michael, my co-founder and I a little bit on, it starts with data management, getting and storing the data and capturing what the firm produces. The second then is governing that, making it permissionable and manageable and to some extent accessible. And then the third pillar is that data enablement. And that's really where we are and where the AI solutions have become, right?

Once customers are ready for that data enablement journey and to actually use that wealth of information, it's then finding the AI use cases, solutions, even niche wrappers that help them do that. So a very good example here of a customer we won back in December. So the way that they're using Pitchly, they're a top 10 investment bank. So they connected,

using our out of the box integration to Intapp Deal Cloud, which brings in the relevant permission data. So again, that data managed and governed. So they were at that level, but they were ready to enable it. Pitchly then designed a prompt in our AI Assist app to then write the marketing description for the Jimstone that met their branding need, their tone, and the information rules and logic to create that piece of content so a person didn't have to.

The description was then introduced to an automated workflow for the banker to approve it and for the legal team to approve it for public display. And then ultimately, the agent also sourced, the pitch, the agent also source the logos corresponding to that tombstone to the company, the counterparty and the client.

Ryan Gerhardy (06:52.138)
So the approval process was both the description, the logos, and then the look and feel. And then once it was approved, it was sent back to DealCloud, the description, that was a new field, the logos, which was new fields, but then also published to their Dan, the digital asset management library for the marketing team, published to the DealCloud CRM for bankers and administrators, and then to PowerPoint add-in, which was upslide for them so that bankers could access it right in their workflow. And frankly, didn't even need to know that Pitchly was doing it.

other than approving in an email the content they created. And so that's the kind of stuff we see really being that next big opportunity of that niche provider. And that's where I think businesses will get the ROI. It's where you can select a AI service, configure it like I described, to be an N of one. That's a very specific workflow to that customer, but extremely valuable.

Jim Merrifield (07:47.215)
Yeah, I can see that. Now I have a follow up question to what you just said. You know, every time a firm or a company selects software, the worry is, you know, what's the lift on, on, on IT. So can you talk a little bit about that? Like say you're a company and you say, you know what, I want to engage pitchley. I want to deploy it. What's the level of, of effort on the IT side to deploy one of these, these agents.

Ryan Gerhardy (07:59.319)
Hmm.

Ryan Gerhardy (08:15.298)
Yeah, and so it varies, which is the worst possible answer, but the truth, it depends if we're accessing information in Excel, that looks pretty low from IT. If IT or InfoSec have decided to use our API to write a custom integration, then it's quite high. With that client, for instance, we use the DealCloud Blueprint. We have it set up already in the Pitchly account. We took templates from their website.

The setup is really done by the administrator of Dealcloud. So IT approves the use case, obviously, but doesn't need to roll their sleeves up other than single sign-on and allowing that InfoSec review and that governance oversight as well.

Jim Merrifield (08:59.088)
Got it. Now it makes sense. I just had to ask the question because that's always the question, right? When you want to purchase a product, you know, I'll go to, I'll be at legal week in about a month from now. And you know, you'll see all these different products on the floor. You'll come back into your strategic meetings and all these great ideas. And, you know, obviously a lot of times, you know, budgets and projects and things are getting really tight for the year.

Ryan Gerhardy (09:02.381)
Yeah.

Ryan Gerhardy (09:24.152)
Yeah.

Jim Merrifield (09:25.904)
So it might just be a strategic conversation for 2025 with implementation in 2026, you know.

Ryan Gerhardy (09:32.588)
Yeah. And I think that's where the world is going. If the data is accessible over an API, we can get to it quickly. That's the question. we build that integration engine and library. Our team can build more of those integrations as needed. But for certain clients, we can jokingly but somewhat intentionally say on a demo, we implemented your account yesterday. Not that we gave you access to the account and...

trained you, but that we set it up from the blueprint. We know the fields you have in your data architecture that we're connecting to because they're out of the box. We can see the templates and the branding and the styles on your website. Sometimes we can even see the bios downloadable and the tombstones and practice overviews. So we can do the whole thing, frankly. So it is a superpower to be able to connect those dots in the background to make a brilliant experience for our customers.

Jim Merrifield (10:27.216)
Yeah, that's exciting. So you're talking about these solutions. Everybody's worried about, there's an innovative side, but there's also an ethical side of development of AI solutions. So how do you ensure that your team develops AI solutions that are both innovative and ethically responsible while maintaining strong IG data management?

Ryan Gerhardy (10:53.07)
Yeah, and think there's, also throw in that kind of ROI. I think that's an after fact concern, but we look at it very much as a, you using, are you giving your customers a choice? So at Pitchly we say, do you want to provide a closed wall experience where we just use the information in your account that's permissioned and approved to be used in Pitchly? Or do you want to use a,

an opportunity to go outside that wall and use things like datasets with chat GPT and use larger LLMs as well.

And so having to able to toggle those things on and off as part of our service is important. So for instance, our Microsoft Copilot agent that works in Teams and soon to work in PowerPoint can request content and data from your organization, from within your walls and nothing else. So when it's not going back and training Copilot, it's simply a one-way feed for Copilot to query our API and get content and put it into that response.

On the other side, I mentioned our AI assist app that can do both. That can go externally and it can use things like sourcing a company industry code, an ASIC code for instance, or a GICS code. It can write a company description. It can get a revenue estimate. It can get employee numbers. It can go to the web and find these things. But it can also internally inside the walls do things like write a tombstone description with these five fields in your Pitchly account.

and it can translate a bio. So you have a French bio as well as an English bio, and it can reproduce that. So they're things that it's not using external data, it's doing it within the account. And then importantly, we wouldn't use any vendors that would use the information of us or our customers to train that LLM. And I think that's really the LLM's job is to license that appropriately for the things they are training with, but then let customers deploy them without that risk at all. And they do, which is good.

Jim Merrifield (12:53.85)
Yeah, it's an important point that you just mentioned that, you know, not using the company's data to train the, the LLM. And, it seems like your product is very flexible with, what you point the solution at, whether externally or internally or both.

Ryan Gerhardy (13:02.307)
Yeah.

Ryan Gerhardy (13:12.686)
And I think we've also learned a lot of customers are still, customers seem to have jumped on this very quickly, or they are still going through how do we approve an AI solution from a vendor? And so they want everything turned off to start with. And with that configurability, their hope is over time, they have the kind of rubric to approve and then can go and do that.

Jim Merrifield (13:34.766)
Yeah, makes perfect sense. So I know Ryan, we talked about a lot here. I'm sure we could we could talk for hours. You definitely intrigued me. I'll probably reaching out to schedule my demo, learn some more and looking forward to possibly meeting you next week or or when this is released this week in New York at Intapp Amplify. But is there anything else you'd like to share with the audience before we let you go?

Ryan Gerhardy (13:47.32)
Yeah.

Ryan Gerhardy (13:56.013)
Yeah.

Ryan Gerhardy (14:02.466)
Yeah, I think my final thought is, I think it's important to have a, a conscious direction and decision on AI at your firm. I don't think the time's gone where you can kind of put your head in the sand and forget about it. I think people will get left behind doing that. And the first steps of that data management, then governance, and now that enablement layer.

I think there's a lot of ways AI can be used even outside of walls with our other partners like Factset or Intapp Premium Data to create content that doesn't need to work with internal data as well.

So there's a lot of ways these can be deployed and the ROI is tremendous. I look at that third kind of, yes, there's the ethical walls and the need to build these solutions in a configurable way for the customer use case. The third side is the AI is, I mean, the ROI is enormous. So 7.8 million assets managed in pitchley. If it takes a minute to create a bio or a minute to create a tombstone, that's 65 years of graphic design time right there, right?

We've seen those numbers 5x from the prior year. And we've seen our data points 20x. So it's only getting bigger and the opportunities there to be taken. So I'd encourage everyone to look at these solutions, run them through many different directions as well. And as we discussed, even having internal technical people help can really give you a leg up. suggesting trials and pilots or different ways to assess this is key to finding your kind N of one fit.

Jim Merrifield (15:38.353)
Yeah, hundred percent. mean, I can tell you it takes internal marketing and BD teams more than a minute to generate a bio for a person, know, probably any company, right? If you're creating a bio page for a new add-on, maybe it's a lawyer or para or whatever, a new business professional, it takes longer than a minute to do that.

Ryan Gerhardy (15:47.148)
Right.

Ryan Gerhardy (16:03.15)
And they're complex workflows, right? Like it's even like one of the surprising things we find is people join in after a couple demos, they say, wait, I don't need to make five tombstones and three bios and choose from 10 different proposal templates. And that's just for one practice area, right? And all of a sudden you realize that not only do you not have to create that, if the deal becomes announced or the template has a change at the template level, all assets are updated.

with our agent, if a logo changes and you want them live logos, they're changed as well. So you get these real big compounding impacts that we help push content, not just to PowerPoint add-ins or dams, but email templates or websites, et cetera. So you really do get that compounding effect and that true autonomy, which is great.

Jim Merrifield (16:50.96)
Yeah, now people might know this. I gotta ask a follow up questions because I know you mentioned this this word tombstone a couple of times and maybe people on the on the you know listening to this understand what you're talking about. But can you explain like what are you referring to when you mentioned tombstone? Because I think right away that I'm going to a cemetery.

Ryan Gerhardy (16:57.452)
Yeah.

Ryan Gerhardy (17:05.699)
Yeah.

Ryan Gerhardy (17:09.134)
Well, you know what, there's a reason the nomenclature is a tombstone. frankly, it's the bane of an existence of me, what, 15 years ago, I was a banker. The reason that a tombstone exists, so they're a transaction tombstone or investment banking style, and it's really to...

memorialize a transaction of securities. And so, and I think it was 1934, the SEC mandated that you have to place a little box describing securities bought and sold in the Wall Street Journal for it to be official. And it became and got this kind of reference as a tombstone or a headstone, if you will, because it's a little portrait box with names and information and a date and a deal value that showed that.

And so this is a very common workflow for transactions, VCs, real estate, et cetera. But yeah, it's very tedious and it's very branded and compliant, right? You have to know that you have their permission to do these things, to have the logos, to have the deal value. It has to be correct and up to date and branded. And it's surprisingly hard to do. The more you get of them, you have to manage this overwhelming library and content.

Jim Merrifield (18:01.38)
Perfect.

Jim Merrifield (18:22.894)
Yeah, and in that context, think was helpful. It was helpful for me, at least. I'm hopeful it's helpful for the audience.

Ryan Gerhardy (18:26.86)
Yeah, and it's a weird like back in the Wall Street Journal, needed to post it because there's no other way to say so-and-so bought so-and-so. But now that still carries through.

Jim Merrifield (18:38.692)
Yeah, great. So, Ryan, again, thanks so much for spending some time here on the podcast on the hot seat with me. This is really enjoyable getting to know you and your company and hopefully have a chance to meet you next week. If not next week, I'm sure you'll be at Legal Week in New York and in March. But again, thanks so much for spending some time on the hot seat with me. If you'd like to...

Ryan Gerhardy (18:48.022)
Absolutely.

Jim Merrifield (19:05.552)
Be a guest on the hot seat, just like Ryan here. All you gotta do is submit your information through our website, InfoGovHotSeat.com And thank you so much. Enjoy the rest of your day.

Ryan Gerhardy (19:13.804)
Yeah, see you Wednesday, Jim, for that demo. Sounds good.

Jim Merrifield (19:17.047)
Awesome. Looking forward to it.