No Priors Ep. 115 | With Glean Founder and CEO Arvind Jain

1. 0:00 – 0:58

   Introduction

   1. EGElad Gil0:00

      (music plays) Hi, listeners. Welcome to No Priors. This week, we're speaking to Arvind Jain, CEO and co-founder of Gleen. Gleen is an AI-powered enterprise search and knowledge management platform, which allows you to not only access all the different internal documents and Slacks and other things that your company may have, but also allows you to enhance workplace productivity by using different applications on top of that. Prior to Gleen, Arvind had a really storied career. He co-founded Rubrik. He was early at Google, worked on search there, amongst other things, and so we're very excited to have him here today. Arvind, welcome to No Priors.

   2. AJArvind Jain0:34

      Thank you for having me.

   3. SGSarah Guo0:35

      So I'm really excited about this. Um, I've known you for years and Elad's known you for maybe 15 more years than that. Um, you're an amazing repeat successful founder with Rubrik and Gleen. Uh, I wanna start by just asking you about search. Um, you've been a search guy, you know, since before it was cool, for a long time when it felt like, not solved, but not as dynamic. Um,

2. 0:58 – 2:05

   How LLMs are changing search

   1. SGSarah Guo0:58

      how broadly has search changed because of LLMs?

   2. AJArvind Jain1:01

      So I've been working on search for almost 30 years now. Long, long time. The paradigm has completely shifted. I think I would say that search had been static for a long time. It was this keyword-based paradigm, like, you know, people ask questions, you find words and try to find them in documents and bring them up, you know, uh, to the users. But LLMs have completely changed it. Like when it's, it has actually, the main thing it has done for search is that it has allowed us to really deeply understand, uh, a question that a user is asking, and similarly, it allows us to very deeply understand what a document is about and you can actually, you know, match people's questions with, you know, the right information conceptually and, and that gives us so much, so much more powers. It's not brittle anymore and I think it's been a, it's- it's- it's been a foundational technology to- to really evolve search into these new experiences that you're seeing these days. You know, where you can go far beyond just surfacing a few links, you know, to, uh, to an end user, to- to actually deeply understand their questions and- and answering them, uh, for them directly using- using the knowledge that you have.

   3. EGElad Gil2:02

      If I remember correctly, Gleen got started in

3. 2:05 – 5:09

   Building out Glean’s platform

   1. EGElad Gil2:05

      the more traditional search world and that as these foundation models and these LLMs have come to the fore, you've really kind of shifted how you think about both the capability set that you provide and how you approach things. Could you tell us a bit more about how you started off building the systems and how that's shifted and then how you've kind of mapped new use cases against it? Because you're now effectively like this really interesting platform that can be used in all sorts of ways in- inside of an organization, around the corpus of information they have. I'd actually love to hear the technology transition. Like how did you think about that? When did it happen?

   2. AJArvind Jain2:31

      Yeah.

   3. EGElad Gil2:31

      I think you really lived through it in a- in a really meaningful way.

   4. AJArvind Jain2:34

      You know, we had good timing, I would say. So let's, you know, started thinking about building Gleen in, uh, late 2018. I started the company early 2019. And so interesting thing is that transformers as a technology had emerged by then. Now the whole world was not talking about it, uh, but in search teams like at Google, you know, we saw the power of embeddings and how it could fundamentally change search. And so- so we had that luxury to actually see this in action. So the version, one of our product actually already used transformers, uh, for semantic, you know, matching. Like, you know, we didn't have these terms, like nobody used to call it vector search, you know, we didn't have that. Like these terms had not been invented yet, uh, or generative AI for that matter. And so like internally, we used to call it embedding search, uh, and it was a core- core technology that we started out with.

   5. EGElad Gil3:23

      So you were super early to it, actually.

   6. AJArvind Jain3:25

      Yeah. Um, and, you know, look, you know, the models at the time were- were not like as powerful as today. Like, you know, we, you know, we started with this, uh, BERT, uh, model that Google had put in open domain, which was trained on all of the- all of the internet's, you know, data and knowledge, and we would then take those models and then for every customer of ours, we'd actually build custom embeddings, you know, on their business content. And then that would sort of power the semantic part of the search. But remember, like in the search, as a technique, there's a lot of focus on embeddings and vector search over the last few years, but that's actually only one part of, uh, building a good search system. Because you think about an enterprise, uh, imagine a company that has been around for a few decades, you know, they have tons and tons of information spread across many, many different systems. A lot of that information has become obsolete, you know, now because it was, you know, written like, you know, many years back. And so when you build a search product, it's not just enough to say that, "Hey, I want to understand if people, you know, somebody's question and I'm going to match it with, um, the right information, sort of semantically or conceptually matches what the user is asking." Well, you've got to solve for other problems too. You've got to actually pick information that's correct today, that is up to date, that has some authority, like, you know, somebody who's an expert on this topic had actually written that document. So you have to do all of those other things too, to actually, uh, truly sort of, you know, pick the right knowledge, um, and bring it back to people. So- so we started with, you know, building the product in- in that shape and form. Um, it was very different product actually. Like nobody had actually search, enterprise search as a problem before. In fact, like, you know, the interesting thing that I remember is that, you know, even though I was coming off of a successful company like, you know, with Rubrik, you know, we had good success, I don't think people really wanted to invest in enterprise search or me, you know, for that matter, because, you know, this problem was not exciting.

4. 5:09 – 8:41

   Why most search companies failed

   1. AJArvind Jain5:09

   2. EGElad Gil5:09

      It was traditionally a very bad problem, right? So there's all these search engines, Fast. I remember when the early Google days was sort of an enterprise search engine, I think based in Norway. Like there's lots of attempts at this.

   3. AJArvind Jain5:18

      Lot of attempts and no successes.

   4. EGElad Gil5:20

      Why do you think it didn't work? Because it- it- it felt like an awful market.

   5. AJArvind Jain5:23

      It was like a graveyard, like, you know, of all these companies that tried to solve the problem and it didn't. Part of it was just that I think search is a hard problem. In an enterprise, like even getting access to all the data that you want to search, it was such a big problem. In the pre-SaaS world, the- there was no way to sort of go into those data centers, figure out where the servers were, where the storage systems were, try to connect with information in them. It was a big, it was a big challenge. So SaaS actually solved that issue. So like search products, like most of them, most of those companies started in the pre-SaaS world, they failed.... uh, 'cause you could just

   6. EGElad Gil5:56

      (music)

   7. AJArvind Jain5:56

      ... couldn't build a turnkey product, but SaaS actually allowed you to, to actually build something, you know, uh, which, which is my insight was that, like, look, you know, the enterprise world has changed. We have these SaaS systems now and SaaS systems don't have versions. Like, everybody... All customers have the same version, you know, they, they're open, they're interoperable. You can actually hit them with APIs and get all the content. I felt that the biggest problem was actually solved, which was that I could actually easily go and bring all the enterprise information and data in one place, uh, and build this unified search system on top. So that was actually a big unlock.

   8. EGElad Gil6:30

      So it was the rise of these connectors and APIs internally. So you're using Google Docs instead of older school systems or you're using Slack or you're using these new tools that now provide you access to the data underlying content.

   9. SGSarah Guo6:41

      You guys must remember Google Search Appliance.

   10. EGElad Gil6:44

       Yeah.

   11. SGSarah Guo6:44

       The idea of, like, I need to slurp your data continuously into a hardware appliance in order to actually do search is ludicrous.

   12. AJArvind Jain6:52

       It was a challenge. Um, the... You know, search as a... And by the way, the origins of Glean is... So at Rubrik, you know, we had this problem. Like, you know, we grew fast. We had lot of information across 300 different SaaS systems and nobody could find anything in the company. And people were complaining about it in our pulse surveys and I, and I was, you know... I always run IT in my startups and so there was a complaint that, you know, that came to me-

   13. SGSarah Guo7:14

       (laughs)

   14. AJArvind Jain7:14

       ... like, I had to solve it, so I tried to buy a search product and I, I realized there's nothing to buy. I mean, that's, that's really the origins of how, how Glean got started as a company and... So that was like, you know, one big issue, like, you know, the... So SaaS made it easy for... To actually connect, you know, your enterprise data and knowledge to a search system, so that actually made it possible for us to, for the very first time, build a turnkey product. Uh, but there were a lot of other advances as well. You know, one is, you know, like, look, you know, businesses have so much information and data. One interesting, you know, fact. So, one of our largest customers, they have more than one billion documents inside their company. Now hear this. You know, when Elad and I, you know, when we were working on search at Google, you know, in 2004, the entire internet was actually one billion documents. You know, there's a massive explosion of content, like, inside businesses, so you have to build scalable systems and you couldn't build, like, a system like that before in the pre-cloud era. I would spend all my time just trying to build that scalable distributed system which, you know, we don't have to anymore because of... Thanks to, you know, all the great cloud technology. And then, of course, Transformers. Like, you know, that's really the real... The big unlock, you know, that we had was that we could actually understand enterprise information more deeply and was very necessary in the enterprise compared to on the web. On the web, even if you don't have good semantic understanding, there is so much that you can learn from people's behavior 'cause, you know, you have a billion people, you know, coming and using your product. In the enterprise, you don't have that luxury, so you have to sort of, like, you know, make, make up for that, you know, lack of signal from users, um, you know, with

5. 8:41 – 10:26

   Out of the box vs. bespoke models

   1. AJArvind Jain8:41

      other techniques, and Transformer is one of them.

   2. SGSarah Guo8:43

      It sounds like you feel a combination of... I'd call it, like, more traditional IR and search techniques and embeddings is, is relevant. Do you think that persists? Like, where would you want bespoke infra or, you know, signals like freshness and authority or, like, how much do models just do in the end?

   3. AJArvind Jain8:58

      Yeah, I mean, I think there's always this thought of that, like, you know, the models will have n- near infinite context windows and you can just give them everything and they can figure things out automatically. But I don't, I don't think, you know, like, we're anywhere close to, you know, that happening. I'll, I'll give you an example. Let's say that models are mimicking human intelligence, right? So they're actually getting more and more capable of, like, you know, how we work, like humans. Uh, but as a human, like, you know, imagine, like, you know, if I were to actually give you... Like, let's say I give you a question and then I say that, "Well here's, you know, here's everything." Like, you know, in a completely non-organized fashion I give you, like, a whole bunch of, like, one million documents and, and, and l- let's imagine you have, you know, the memory powers and speed, but it still just feels like, you know, a very complicated thing. Like, it's very hard to make sense of information that is, for example, being given to you out of order. Like, can I give you one document that is something from today, something from four months back, something from three years, then something again from two days back? If I give you, like, you know, information in a, in a, in a manner where, you know, where it's sort of not organized in any shape or form, then as a human you're gonna have a lot of difficulty reasoning over it. So do we think about the models the same way? If there is a good amount of, uh, work that you have to do and present the information, uh, to the model in some s- you know, in some organized fashion, that's when they're gonna actually do a much better job reading that information, reasoning over it and giving you the answers. And sure, like, you know, you can actually give them more and more over time, but still, you know, it matters, like,

6. 10:26 – 15:34

   Creating apps on top of internal knowledge

   1. AJArvind Jain10:26

      you know, how, how you provide them with the right information.

   2. EGElad Gil10:28

      Now that you have this sort of corpus of information, right? You've basically aggregated all the internal documents of a company, which in itself is incredibly useful just for search, but you've also gone the route of, like, enabling applications to be built on top of it in different ways. Could you talk a bit about that and what are some of the common use cases that you're seeing?

   3. AJArvind Jain10:43

      So we started with, you know, this vision of building a Google in your work life, um, but then as models got better, developed, like, these reasoning and generation capabilities. So first, like, you know, it changed our product and, uh, like, our new product, like, Glean Assistant, you know, it sort of looks and feels more like ChatGPT. Um, so instead of, like, you know, me going asking questions and seeing, you know, a bunch of links coming back to me, you know, now of course you converse with Glean, you ask questions and it works just like ChatGPT. You come and ask a question, it's going to actually take all of the world's knowledge and also additionally, you know, it's gonna take all of your internal company's, you know, data and knowledge and use that in a safe and secure manner, like knowing who you are and what information you can really use within the company, uh, to answer questions back for you. So, so that's sort of, like, the first, um, first progression in terms of our product. Like, you know, we evolved from being a Google to, to, you know, something that looks more like ChatGPT, a more powerful version of ChatGPT inside your company. As you build that, this Glean Assistant actually... You can th- think of it more like a personal assistant that you're actually giving to every employee in your company. It's a tool, you know, it's your sidekick, you know, it's always available to help you with whatever questions or tasks you have. It's gonna use all of your company's context and data to help you with, you know, with, with your work. But, you know, businesses are actually...... a lot more, but more interested not in that, but in actually thinking about how they can transform their company with AI or they can take specific business processes, you know, where they're spending a lot of money, uh, and how, how do they bring automation in that with AI. So we've been, we've been asked like before agents became, uh, you know, the talk of the, like, you know, of, of the day and, like, now everybody's of course building agents. Uh, but early last year when agents had not yet taken off, people were asking us for that, "Hey, give, you know, we need, we need to build more curated applications, you know, using this data platform that you have." So as an example, HR teams, you know, would come to us and say that, "Look, we love Glean Assistant." People come in there, ask questions about, you know, benefits and, you know, PTO and vacation policies and whatnot, and, and it work, it works great, but sometimes it uses, you know, content that's not authorized or blessed by us. And if somebody's coming and asking questions on people-related topics, we want Glean to only use, you know, the curated content that, you know, our people team has created, uh, and we want it to behave in a particular way, particular tone and all of that. So, so that's a, that was a request that we started to get last year that, like, you know, can we create, create more specific curated experiences, you know, function by function for, you know, for different use cases. So we started to build that and we were not calling them agents, we were calling them apps. Uh, now of course, like, you know, the people think of, you know, them as more as agents because it's no longer just, you know, asking questions and getting answers. But you want, you know, these, these specific functional experiences to, to actually replace a business process, which also involves doing some work for, um, you know, not just answering questions, but actually doing some work in those systems.

   4. SGSarah Guo13:44

      Arvin, when you talked, um, about, you know, access to the, the right data with the right authority and also... Like, it really begs the question of like access control-

   5. AJArvind Jain13:53

      Mm-hmm.

   6. SGSarah Guo13:53

      ... right, in a, a platform like Glean when you have all this unstructured data, um, uh, this seems much more complicated. What is, like, your overall stance or how, how you think this is going to work in the future?

   7. AJArvind Jain14:05

      Yeah. Well, so look, enterprise information in some sense, you know, it's governed and it's protected. Um, you, like most of the knowledge, I should say like 90% of the knowledge inside the company is private in some shape or form inside the, within your company. Like, you'll have a document that maybe is private to you or, or you share with a few other people, but that's the nature of, um, you know, enterprise knowledge. Uh, that's, that's the fundamental sort of, uh, way, like, it works. And you can't take, you can't actually build like, for example, a model inside your enterprise and dump all of your internal company's data and knowledge into it and then make that model available to everybody in the company because if you do that, you're leaking information, like, you know, inside a company, you're letting somebody in the engineering team, you know, see sensitive stuff, you know, which probably only HR team should be able to see as an example. So any AI experiences that you build inside the company has, it has to think, you know, about security and governance and permissions, like, you know, at a fundamental level. And that's what we do in Glean. So when we connect with all these different systems, um, you know, inside your enterprise, we, you know... If, if we index, you know, a particular document from Google Drive or a conversation from Slack, we also keep track of, you know, who are the users can actually access that information. And this is fundamental, like any access to data that's going to happen through our platform is going to actually match, like, you know, the users have to be signed in and we will actually only let them use information that they have permissions for. And, and this is, this is important, uh, as a problem to solve, like, you know, unless if you have infrastructure like

7. 15:34 – 19:11

   User behaviors & insights

   1. AJArvind Jain15:34

      that, you cannot roll out AI safely inside your enterprise.

   2. SGSarah Guo15:37

      I learned a lot from people who work on search, especially like search with any sort of scale because you get all sorts of weird user behavior. Um, and so related to your i- idea of, you know, us with our personal assistant team, what are some behaviors you see from end users in terms of how they're using Glean or AI in general that you think we should just do more of, right? Like, I, you know, I, I'm always very surprised when I learn from Google people about just, like, the behaviors around navigational search and how many are one-word queries or what the popular queries are and those sorts of patterns. And so I'm sure you, you see like Glean and AI superusers.

   3. AJArvind Jain16:10

      One of the biggest surprises for me, um, I always felt that, you know, we are building such an intuitive product, you know, the, it's like it's literally, there's no UI, you know-

   4. SGSarah Guo16:20

      Yeah.

   5. AJArvind Jain16:20

      ... there's one box and you ask question, you put in a search and, and, and what's the big deal? Like, why do you have to learn how to use this? Um, and, and we realized that as we added more and more of these natural language capabilities and, you know, the ability for you to actually ask a really long question, like, you know, paragraph long, you know, uh, set of instructions that you're giving to us. Um, and we realized that people won't do it. Like, you know, I think everybody has been trained over the last 20 years to actually type in, you know, one or two keywords, like Google has sort of taught us, you know, you know, on what search can do. So with search, we never had a problem. Like, you know, launch a product with, like, immediate high usage. Uh, nobody was confused like how to use the product. With, with Assistant, people didn't know what to do with it. Some people with, you know, m- more curiosity and they will, they will ask all kinds of, you know, questions that we couldn't actually answer. Um, for example, somebody says that, "Hey, what should I do with my life?" So we, so I think, but anyways, coming back to this, that, that was one of the key learnings is that AI is, is actually s- very un-intuitive. For most people, you have to actually really ex- expose to them these capabilities in a, in a sort of a incremental fashion. You know, like some things, you know, which sort of are more meaningful to their day-to-day work. For example, if I'm an engineer, um, like, you know, prompt the user sometimes that like, "Look, you can actually learn about a new piece of technology. Like, I can actually give, you know, create a two-page tutorial for you right now." Um, and, and you sort of have to understand like, you know, what people's, you know, p- people, like, you know, what their core work is and, and then you have to actually give them these, you know, sort of prompts, like prompts for them to sort of start experimenting and get excited about like trying something out with AI. One thing in fact, which I would also add here is...Like, a lot of time, you know, with AI, businesses are excited. Like, you know, they have a lot of dollars to spend on AI and... But they're all, like, also asking for an ROI. That, well, like, you know, I'm going to make all this, you know, investment, what are the returns? Like, what are the efficiency gains, you know, that I'm going to be getting or what are the top line, you know, improvements that I can make to my business? Uh, there's a lot of focus on that. And I think one thing that often gets overlooked is education. 'Cause you know the world is changing. Imagine like, you know, three years from now, you wake up, you're the CEO of a large enterprise. What do you want to see in your workforce? You actually want to see people like who are trained and are AI first. Like, you know, they're experts, they know how to leverage... The strengths of AI 'cause this is a difficult technology. Like, you know, it's not perfect, it's not easy. Uh, it makes mistakes, it hallucinates, but yet it's powerful. And if you become an expert, um, you can get a lot done with it. That has to be... Like the objective today is like, you know, with like... Like as leaders, think about AI, how do you sort of give people tools that sort of motivate and motivate them

8. 19:11 – 21:51

   Unique challenges of building Glean

   1. AJArvind Jain19:11

      to bring AI in their day-to-day work?

   2. EGElad Gil19:13

      You had an amazing career between being early at Google, starting Rubrik, now starting Glean and running it. What was unexpected about doing Glean? 'Cause you'd gotten to so much scale, you'd done such amazing things in, in the context of Rubrik. What was hard or unexpected or just very different about Glean that you, that you didn't anticipate?

   3. AJArvind Jain19:31

      From a product side, the... One of the most interesting things for me was, like how hard was it to actually, uh, roll the product out to our customers? We had a very different journey, um, like in, you know, Rubrik compared to Glean. Like in Rubrik we're an established market, like there were budges, there were dollars and you had to actually replace an old technology with a new technology. Um, here, we were in a market where, uh, we had no, uh, budgets, there was no concept of buying a search product in the enterprise and, and everybody thought that, yeah, like there's an important problem but I'd like... You know, it's not a line item in my business priorities. You know, it's, it's a, it's a vitamin, it's a painkiller. People are living without it. Well, yeah, that's, that's true. I mean, you live without something that you don't have, like, you know, that's by definition, you know (laughs) you know, true. So we, so we had a lot of challenge. Like we had to do a lot of evangelism to actually get the right... Like, you know, um, folks, like, you know, who wanted to be, um, the innovators, like for them to actually make that bold call and, and actually buy a product that's, you know, they're not used to buying. So that's, that's sort of first part of it. Like, you know, you have to create the market for this, which actually was difficult. And second, it was actually very interesting one, is, you know, we, we... Our product was actually working well, like, you know, it was doing good search or that, you know, letting people find things but then we started to hear from businesses that, "Oh, I'm scared of good search. I don't want, I don't want a good search-

   4. NANarrator20:53

      Mm-hmm.

   5. AJArvind Jain20:53

      ... product in my company because I have all these governance gaps. I have like, you know, sensitive information all over, all over the place and, you know, now people are discovering these things." Then we launched, like, you know, for example, you know, people found like salaries of other people, you know, those... Like in one of our customers somebody found a sensitive M&A doc that was, you know, for something that was, you know, not yet happened. Uh, and you start... Like, so people like you actually were very, very scared of actually having good search. So we had to actually... Like that was, that was interesting challenge. We did some good work, we were doing it safely and securely but, you know, you don't have good governance and now, like, you know, we don't... We can't sell because our product is so good.

   6. EGElad Gil21:30

      It seems like LLM should be able to help with that, right? In terms of classifying documents and surfacing, "Hey, this one may be sensitive. Do you want to secure it?" Et cetera.

   7. AJArvind Jain21:38

      Yeah. So in fact that, that's e- exactly right. Like, no, so we actually were forced to build that. We were forced to actually go and above and beyond respecting permissions in individual systems to knowing who you are, what you're asking, like usually

9. 21:51 – 25:00

   Product-led growth vs. enterprise sales

   1. AJArvind Jain21:51

      have the right to ask the question. Or, like, you know, when the information comes back, like does this things, you know, feel, you know, safe enough for us to show it to you? So we actually in fact... Like, you know, in that sense, you know, we actually end up, ended up becoming a security product. Like a lot of companies actually buy us to fix governance in their sort of, you know, data and systems and become AI ready. Like AI ready for the Glean search product, uh, the Glean assistant, but also for all the other AI products that you can buy inside the enterprise. So that was, so that was actually a very interesting journey. But then for personally on, you know, for me, um, you know, at, at Rubrik, you know, I, I didn't actually, I wasn't the s- I wasn't the CEO. Uh, I ran R&D as one of the founders of the company. And here I had to actually learn how to become a CEO and I don't think I've learned it yet. And like, you know, that's a, that's a constant, you know, challenge and like, you know, learnings, you know, that I go through. Because fundamentally, like, you know, I'm still an engineer. Everything I do, like, you know, like, you know, that's the mindset that I have. So, so growing, growing, you know, out of that into like, you know, being able to run a large business, you know, that's a, that's a personal transformation that I'm going through.

   2. EGElad Gil22:52

      One thing that I think is striking is that, um, from a GoToMarket perspective, you all are really focused on big enterprises, right? And you mentioned some of these enterprise data needs. A lot of people always just want to do PLG, and you've really done sort of the top down sale and it's been incredibly successful and you've done it twice now, right? 'Cause Rubrik was-

   3. AJArvind Jain23:07

      Yeah.

   4. EGElad Gil23:07

      ... largely that as well. Could you talk a little bit more about when it makes sense to do big direct enterprise deals versus a PLG motion? And how you think about that as you build businesses? 'Cause I think it's very differentiated and most people just can't pull that off. So I was, I'm curious about how you think about when to do it and then how to do it?

   5. AJArvind Jain23:22

      Just to be candid, Glean, the... When we started, I mean, my dream was to do PLG. I'm an engineer and I wanted the company to have engineers and then product should sell itself, you know, on the web. Who doesn't want that? It was, it was something that, you know, was a desire for us. Uh, but the problem is like, you know, with our product, the... It is by definition a, a company-wide product. Like it's not like, you know, we cannot offer the product to one individual inside a company. Even one person, you know, their, their search needs require us to actually search over all, the entire company's information for them. So it's a, it's, it's expensive. You have to actually index, you know, all of your company's data and knowledge. And so we never had that, uh, concept so that, you know, we could make it available to one or two or ten people inside the company. So we were sort of forced just structurally to actually build in that fashion where it is, it is, you know, like enterprise. It is, um... Like, you know, we roll the product out company-wide, you know, to every employee. Um, that, that's what makes it cost-effective. But like, you know, coming back to your question, the standard approach I think that companies prefer now is that, like, they think of PLG as basically Reach and as a funnel you sort of nurture and expand using, you know, enterprise sales motion. So, the right recipe for me, like, you know, if I had a choice, I would, I would actually start both the mo- motions simultaneously. Like, I won't actually say that, "Look, you know, for the first three years I'm, I'm gonna actually focus, you know, uh, on just doing PLG and then bring enterprise sales later." 'Cause you're actually leaving a lot, you're leaving a lot on the table. Uh, timing, timing matters always and, and so you have to sort of like start the, start, start the motions I think at the same time.

   6. SGSarah Guo24:54

      Arvin, one thing that we have talked about that I feel like must have been, um, I mean, hard, the priors

10. 25:00 – 27:08

    Succeeding in traditionally bad markets

    1. SGSarah Guo25:00

       on this market were not great, right? And we talked a little bit through the rationale of like, you know, you feeling like you, you really saw the problem internally anyway and understand that there are these, uh, sort of architectural, foundational things that had changed in terms of movement to SaaS and API-based integrations and such. But still, I think it's a really good question of advice for founders or maybe people joining startups like, uh, when should you agree with the priors on like something is a bad market or, or how should you think about that question?

    2. AJArvind Jain25:31

       So I'll share a few things on this. Um, number one, I think as engineers like, you know, there's, there are... first of all, there are always doubts like, you know, the more you look at priors, the more you're gonna actually... likely you're gonna actually ultimately kill your own idea. There is a lot like, you know, sometimes...

    3. EGElad Gil25:45

       Everything has been tried, yeah.

    4. AJArvind Jain25:47

       Yeah, everything has been tried like some, you know, a lot of things have failed and, and I think there are... again, like for any given idea like, you know, there are 10 reasons why it won't work, um, like as you start to go into details. Sometimes like a more simpler approach is helpful, you know, which is, uh, well, there's a problem like, you know, you talk to people, they have and they feel this pain and... which clearly means that nobody is actually yet solving, you know, that... because the pain exists. And so then don't go into details anymore, just do it. Things will just get figured out over time. So like, at least, you know, for me like, it was actually unusual for me like I'm, I'm an engineer by training myself and I'm, I'm like, you know, I'm naturally trained to question and like there's a lot of self-doubt in my mind. So I don't know what happened to me when we started Glean because you know there were all, all these people saying like, "No, not do it," and, and somehow they couldn't like, you know, they couldn't actually discourage me. Like, you know, I just felt that this was an exciting problem, um, by, you know, I was, I was... I knew everybody in the world, you know, has this issue like, you know, even at Google, like it was a big joke, you know, always we had internally like, you know, we're... All of us were spending all of our time making it easy for people to find things, but not us internally at Google. Uh, it's super hard to find anything inside the company. So, so I think I, I so- somehow like found that conviction. I was sort of being lazy, not willing to go into the details and like, like look at all those priors and just like, you know, just, just do it, just solve it. I mean, I think that's, that's

11. 27:08 – 31:33

    What Glean is excited to build next

    1. AJArvind Jain27:08

       what I think like worked for us in thi- this particular case.

    2. EGElad Gil27:11

       I feel like Glean had like three big components to it that all came together that you mentioned earlier, right? There was the need that you identified just as, uh, somebody running IT for your own company and to your point, it goes back to Google that this was a need. And every company that I've talked to has always wanted to build search and directories and all this stuff. The second thing is this rise of connectors and APIs in the context of existing enterprise software that everybody's using so you can extract the data more easily. And the third thing was the big shift in terms of the underlying technology, right? The shift in terms of what is capable of a search, these foundation models, these embeddings, etc. Given, um, the latter two, are there other big opportunities that Glean isn't gonna work on that you've kind of identified as really interesting areas that suddenly are tractable again?

    3. AJArvind Jain27:51

       I think for us right now, the focus remains on the two core products that we have. So we... you know, our... the way we think about our company is that we have this really powerful end user, uh, AI, you know, assistant that helps every person like, you know, work differently in the future. And then we have this agent platform that you can use to actually bring, you know, AI, inject AI into, you know, every one of your business process, make them better, make them more efficient. And, and I think we are... we're making big promises on both to our customers. The way I describe and pitch our product to our customers is the following. You know, come to Glean, ask any questions or give it any task, Glean will use all of the world's knowledge and all of your internal company's data and knowledge in a safe and secure way and answer those questions for you or complete those tasks. But actually, I just promise to you that Glean does everything, you don't have to work anymore. We're long, long ways from actually even solving, you know, that, you know, the pitch that I just mentioned to you. Like, you know, I think we have to understand, um, knowledge properly. We have to like, you know, pick, you know, the right, you know, the correct information, throw away the old information. We have... you know, there's so many challenges, there are so many issues. Uh, people talk about hallucinations as a big problem with AI models. You know, we feel like, you know, a bigger problem for us is not about hallucinations, it's about like, you know, most of the times you can't even, you know, find the right information, sometimes it's not there. People are asking questions, but nobody wrote it down. Uh, sometimes, you know, we are not able to actually do the needle in the haystack, you know, we pick the wrong thing. And so there are like a lot of challenges and I think we, we will be working on this problem for a long, long time. And I don't see us having any need, by the way, like, you know, of wanting to do something different, like, you know, like just solving this one problem itself is a big, is a big, big, um, success. So, so we're gonna stay focused on these two, you know, these two products. Um, but then they're also like, you know, talk to, talk to you about... a little bit about the vision for the future. So I think the way we all work is sort of an accepted that AI is gonna change everything. AI is gonna change how people work, AI is... AI are going to actually change how businesses actually even look and feel, you know, what kind of, you know, workforce you have in the future. And one thing that's gonna fundamentally happen is that each one of us is gonna have this amazing team, um, of, you know, call it assistants, coworkers, coaches that are totally personal to you. And, you know, you're always surrounded by that team and this team knows everything about you, your work life, um, what you need to do today, and this proactively helps you, does 90% of your work for you. And also like, you know, help you get better, like, you know, at your, at your, you know, like upskill you, um, be your coach. Um, and this... and that's the world that we want to be living in. Like today, you know, there are some people who have... who alre- who already live in that world. Like, you know, for example, being a CEO, you get the luxury to actually have all of that. You have assistants, your chief of staffs, you have an exec team, you have a coach. Uh, but in the future, that's going to be something that all of us are going to have like, you know, regardless of how senior we are, you know, may- maybe a new grad joining the work, joining the workforce. That's what we are trying to actually, um, go and solve for. We're trying to actually build that amazing personal team around every individual, um, that's going to make us all 10Xor. And that's just a natural extension of like just keep evolving our Glean assistant product to make it better and better over time.

    4. EGElad Gil31:06

       Yeah. Arvin, thanks so much for joining us today.

    5. AJArvind Jain31:08

       Yeah, that's excellent. Yeah. Fun, fun questions.

    6. EGElad Gil31:11

       It's always nice to see you.

    7. AJArvind Jain31:12

       Yeah. Likewise.

    8. SGSarah Guo31:15

       (instrumental music plays) Find us on Twitter @nopriorspod. Subscribe to our YouTube channel if you wanna see our faces. Follow the show on Apple Podcasts, Spotify, or wherever you listen. That way you get a new episode every week. And sign up for emails or find transcripts for every episode at no-priors.com.

Episode duration: 31:34