The Exchanges

Q What are some of the use cases? If I'm a, you know, an enterprise potential customer today, what do I use Rider for? And then we'll get into the, the, how it works behind the scenes.

A So if you are a major investment bank, let's say you are the top investment bank in America, um, you are using us for everything from earnings calls, summaries, to deep dives on specific companies and sectors. Um, you're using us for search over merger proxies and M&A docs. If you are a Salesforce, they're a major customer. Um, you're using us for dozens of different kind of custom applications plugged right into Slack to review content for compliance, to automatically, uh, rewrite for SEO, to produce marketing and email. Um, if you are a, uh, insurance company, CSAA, uh, uses us to build Uh, knowledge assistance for agents who are answering the phone. So our, our focus area is, and the verticals that we focus on are financial services, healthcare, and retail CPG. And, you know, the, the, the types of use cases are, um, everything from your mission critical, like reviewing, uh, insurance policies and doing claim adjudication on them, um, to helping Salespeople sell more. Now, when we are in call centers and support, it's not, you know, chatbots for your website or helping exchange mediums into smalls. Like, there's lots of companies doing that. It is really in supporting, ah, a high-end kind of knowledge task. So, if you are, um, a CPG company and somebody calls into the call center, To ask whether, you know, there's phthalates in the, uh, Avino product, right? You are, as some…

AI assessment note: “you are using us for everything from earnings calls, summaries, to deep dives”

Q So graph-based rag is sort of the cool term du jour, the cool approach du jour compared to vector-based. Can you maybe compare and contrast?

A In a vector-based approach, what you're doing, and I'm, I'm grossly oversimplifying here. Let's say we are looking at, um, a insurance policy, and in your head, imagine that you've got an insurance policy that's actually scanned, and you've got tables, and you've got, you know, two or three columns on, on every page, and there's tables, you know, nested within, um, that insurance policy. That is like good data when you're talking to the enterprise, right? Um, and if you are taking a vector-based approach, um, to answer a question like, um, you know, I, um, uh, I'm talking to a client, and here is their policy, um, and they want to know, you know, what's the coverage maximum if, um, you know, the trailer gets rear-ended, or whatever. When you are doing a, a vector-based approach, you're going through, and you are, you've chunked Right, that policy, and, you know, against the prompt, you are trying to find the most relevant chunks, right, in that, in that policy, which means it does really badly, right, with tables, because you've lost, you've flattened out the document, you have no context for, like, where this nested table is in an insurance policy. It does really poorly with numbers, um, and it does really poorly when you've got lots of policies that are all talking about Trailer is getting rear-ended. And so, and there's a ton of techniques to be able to get, you know, the si…

AI assessment note: “Compare that to a graph-based approach, um, where we've automatically created the nodes”

Q to a two billion dollar valuation. Um, in which case, uh, I may or may not congratulate you. This is the information I have. And, um, in terms of history, before you started Ryder, you founded a company called, uh, Cordoba, that was doing, uh, machine learning, and, uh, I believe a focus on machine translation, sort of a grammarly kind of, uh, focus. Can you talk about that story?

A It's been, it's been really one continuous journey for me and Wasim, truly. And, uh, we have been in the realm of automating language for the entirety of the 10 years we have been working together. So in the, the first company, Cordoba, um, we started with statistical machine translation. So this was years ago. And we started using, uh, transformers and map business encoder decoders at the time. Um, really to solve NLP problems associated with, um, with translation. But it was, uh, very easy to see, you know, how powerful, um, this tooling was. And, um, we, we, in a way felt like, uh, after years of working on a, um, a very mission-driven business around, um, you know, making the language you were born speaking just a non-issue, uh, especially in the workplace. Uh, to, to go work on source language really felt like abandoning, um, that initial mission, but the technology was just too exciting. So in a lot of ways, the story of Ryder is the story of the Transformer, and it was in the first business that, um, we really started, um, to, to use the technology. And Rider officially started in, in, in, in, in, in, in, in, in, in, in, in, in, in, in, in, in, in, in, in, in, in, in, in, in, in, in, in, in, Really just the EPD team and that core team. So our core team has been working together for years, um, you know, six years, seven years, some of our oldest employees across two compa…

AI assessment note: “So in the, the first company, Cordoba, um, we started with statistical machine translation.”

Q Okay. So how does, how does that work?

A Yeah. So, um, there are a lot of different guardrails in, in Rider. Um, and the, the approach is different depending on what the guardrail is. So if, um, if it is PII, you're literally going into writer and you click a button under compliance settings that says no PII, right? It's that simple. And unlike, you know, bedrock, like this is not just like a simple reject filter because that doesn't work, right? You really need, again, an LLM based kind of understanding of like, what is PII? Um, On brand, right? Um, we're doing post-processing to introduce a, a rewrite, an LLM-based rewrite based on a brand's guidelines. Um, and that is also, you know, think of it as a fine-tuned LLM, right, that rewrites everything that comes out. And again, that's like a multi-select, right? I'm a pharma company, and I have 21 brands, and I'm, you know, completely standardizing How I build call scripts for my field sales rep. So you're going up to see an oncologist. You literally have got this objection handling script for a drug. This is stuff that we work on. Um, but it's gotta sound like, you know, in addition to all of the compliance, right, around that, it's also gotta sound like, you know, whatever the drug is. And so coming out of Rider, you're able to put that as a guardrail, right, to make sure that whatever is produced aligns with, with those, um, with those guidelines. Um, Specific types…

AI assessment note: “the approach is different depending on what the guardrail is”

Q What is your sense of where we are in the adoption curve for enterprises? What, uh, what is the positive? What is the negative?

A I, I think we're very, very early, which is of course so exciting because we see so much opportunity already. Um, right now we can go to the market of truly early adopters and talk tech all day long, right? Literally the CEO knows what we're talking about when we say graph-based versus vega, and I'm not like, this is fortune 10, the CEO knows. Um, the, the mainstream market, we're gonna have to skew all of this stuff up as we go to the mainstream. Market. And so, I think that's very exciting for, for startups. I see a lot of startups go to the enterprise, and you can definitely get demos. I mean, literally, I was talking to a CIO yesterday. Fortune 50 had just taken personally a demo from an eight-person startup, because people are so curious. They, like, want to know. They need to be the smartest. Um, but you can't ask the company to, you know, like, build the last 50% of your product. You have to, the stakes are really, really high. Um, because everybody else is there, right? The consultancies and the strategic advisory firms and every hyperscaler can literally throw more innovation dollars at the enterprise than you will raise in three years from venture capital, right? So stakes are very high. Um, and I, I do think that we, we've been very fortunate that we've just been enterprise from day one and really got to kind of build and Like total obscurity, um, with the, the enter…

AI assessment note: “I think we're very, very early, which is of course so exciting”

Q The AI studio, is that where people can build their applications? Is that what it is?

A Yes. And, you know, there the, the insight, um, really came first from, and this was starting not that long ago. I mean, just this is space has moved so, so fast, but starting about two and a half years ago, um, and we started in marketing, but for highly compliant, um, industries, folks started to need just way more customization. Like, Hey writer, you know, the stuff coming out of your generate model, it's fine, but I really got to kind of squint to think that it's even good, right? I'm not using it. And so for us to really get amazing required a ton of scaffolding around the LLM, right? And like prompting is not even the right word for it, right? Like we are really, let's say you are, um, uh, a Franklin Templeton or a Vanguard and you're producing Market commentaries based on fund tear sheets. Um, we need to, like, read the charts, um, we need to understand the graphics, we need to understand the commentary, and then compare cross font to be able to write, you know, the, the commentary on it. And so, you know, it started very early for us that, um, the, the type of content that passed muster, right, and got to production grade and really wowed people, Um, just required a ton of business logic around, um, you know, what, what you ask the, the model to consume and, and produce. And, um, you know, this was L'Oreal about a year ago, you know, literally like a 180 apps, and we ne…

AI assessment note: “Yes. And, you know, there the, the insight, um, really came first from”

Q What are you excited about, I guess, in AI in terms of like where things are going, whether that's, I don't know, additional modalities or like new models coming online. What, um, what would you find yourself naturally getting curious and excited about?

A We got some fun, exciting news today from a major analyst, um, that were included in, um, another big quadrant around, uh, agentic, um, for, for this next quarter. And, you know, I don't like that word because there's millions of people who are that job, you know, they're agents for all sorts of things from insurance to sales, et cetera. Um, We, for, you know, that functionality, um, are, are using autonomous action as, as the, the word, and I'm really excited for what's possible. And you look at a lot of, um, the, the demos and what folks are trying to do today, and, you know, it really just amplifies the weaknesses of, of LLMs. Our approach on autonomous action is really to start with the most knowledge intensive, the most important nodes of a workflow. And then, you know, autonomous action, um, can, can really do some incredible work in, in moving from system to system and, and really being able to, to reason in novel situations to take action autonomously and proactively with all of the right observability and controls, of course. Um, but especially when we have seen, you know, our early customers, so customers who are early on our autonomous action, um, functionality, but have been customers for two, three years, like just, Get to square 10 because they've gotten to square one. It's so, so exciting. Um, and so I'm most excited about that. I do think, you know, 20, 25 is, i…

AI assessment note: “and so I'm most excited about that.”

Q You got the AI studio. You got the knowledge graph. You have the guardrails. So let, let's start with the graph that you just mentioned. What is it? What does it do?

A For so many use cases, let's say you are, um, a major CPG company. This is a lot of, you know, what we do at L'Oreal, at Kenview, et cetera, and regulatory affairs and compliance are, are using Rider to go through, um, you know, federal guidelines, federal, federal regulations, stuff that's getting updated in, in real time, um, to produce arguments for why, um, you know, uh, we've got, The, uh, packaging that we do or why we can make the claims that, that we do. Um, when you are doing kind of this desktop based research using LLMs both to kind of go in and digest the regulations and then actually build and write the report. These are great use cases for us that produce a ton of value. Um, you are not going to be asking the LLM directly those questions, right? You are going to be using an app, um, that uses A RAG pipeline to be able to really get, um, that, that data into the LLM. So we do have domain specific LLMs, um, but for like a use case like that, you need to set up, set up RAG. And a lot of times, and we almost always, you know, we're, we're not as big as Microsoft or OpenAI yet, right? So, you know, we're the brand that is coming in and the company that's coming in when there's already a DIY project, right? Like these companies have all been Trying to crack the nut for the last couple of years. There is something that just hasn't scaled when we are talking to customers.…

AI assessment note: “These are great use cases for us that produce a ton of value.”