Appian Corp Corporate Analyst Meeting Transcript

The following is excerpted from the question-and-answer section of the transcript.

Question: Jacob Roberge - William Blair & Company L.L.C., Research Division - Analyst : This is Jacob Roberge with William Blair. You talked about this morning during the keynote, a lot about private AI, and you mentioned it a few times here. What's differentiated about that? And is the data set that your clients are giving you large enough to really train these models off of when you're going up against those public AI models that might have a much, much larger data set because the data is endless versus your private AI model. REFINITIV STREETEVENTS | www.refinitiv.com | Contact Us consent of Refinitiv. 'Refinitiv' and the Refinitiv logo are registered trademarks of Refinitiv and its affiliated companies. MAY 02, 2023 / 8:30PM, APPN.OQ - Appian Corp Corporate Analyst Meeting Malcolm Ross Well, the -- it's not -- to understand private AI, it's not our private AI model. It's the expectation that customers are going to want private AI. So the difference between a public AI service, which exactly as you say, has a huge data model is that it also is a huge generic AI model. It's not representative of that company's unique business. So that company's unique business, they might have millions of customer communications every single year, which never go into that private AI model because of there private communications with their customers. So how can they create trained models on their private data such as e-mail communications, while making sure that they're not just sharing that with a public service. So our anticipation is really that while these public AI services do become useful at a generic sense, a great example is natural language understanding because we all hear, I think, speak English. So having an AI model that also understands English is useful for everyone. But how many of you are medical experts? That's proprietary knowledge in a specific person. So that's kind of an equivalent of unique knowledge that enterprise has and how can they create those custom AI models on their specific knowledge, which is what we call private AI. And so private AI is really a motivation by customers to be able to differentiate using AI but not relying on public services by having low-code tools to build those proprietary AI models. Does that make sense? So other questions? To you, Suvajit. PRESENTATION

Question: Jacob Roberge - William Blair & Company L.L.C., Research Division - Analyst : This is Jacob Roberge with William Blair. So just a little bit of a different spin on that question, but we've talked about earlier today this and then process automation and really in the optimization around that and how Appian has built low code, it's brought together process mining. It's brought together RPA. Do customers that you speak to understand the value prop to combining all of these into one platform and adopting the full platform versus piecing together the puzzle with the UiPath or the different solutions that are out there. Unidentified Company Representative Absolutely. Customers do understand. If you think about a lot of tools came out, which are purely RPA for quite some time, there was tons of bots built beyond a point, it was a governance nightmare. So customers struggle and then you move into a low code for certain use cases. You've got mining that is, again, a differentiated part. So when you think about these different investments, customers definitely struggle and beyond a point, governance nightmare and dealing with relationships and stitching these 3 different pieces together. So it is always a good value proposition when you can solve a particular use case, when you can solve a particular business problem, think through from a mining standpoint, from a process reimagination standpoint, designing the process and then, of course, plugging in the last mile through RPA. So customers definitely see that differentiation.