Transcription
The technology stack is evolving at a pace we've never seen before. By default, I have this mindset: if you built something last year, it's already obsolete today. There must be a new and better way to do the same thing. And if not, it's simply a lack of imagination. Startups in general are very hard work, but there's also a lot of joy in it. You work with people who are truly motivated and moving towards a common goal. And you really like that feeling. We are in this race together. We are going to build a great product, and we will win. We will attract customers. Working in the industry doesn't make things easier, because it's perhaps the most fiercely competitive industry out there. So much change, so much competition, that you don't feel like you can even get off the treadmill to grab a drink of water. That's what it feels like to work at her company. But we know why we're doing this. Welcome to Grid. I'm Jubin, a partner at Kleiner Perkins. This is a show where we go beyond the glossy success stories and break down the personal and professional challenges of building companies that make history. Today, we have with us Harvin Jain, founder of Glean, an enterprise assistant and workspace search platform. Glean, by the way, started in the basement of Kleiner Perkins. Arvin is in our show for the second time. It's incredible to see the progress this business has made in just a few years. How big is the company now? How many people? We've crossed the mark of a thousand. No way. Yes, honestly, I was surprised myself. And to be clear, it wasn't a feeling of celebration, it was a feeling of panic when I saw that we had exceeded a thousand employees, because that's a very large number. And why panic? What is it about that number specifically? Like, are you worried that the organization has become bloated or that you're turning into a big company? Well, look, it's often said now that a company with a billion in revenue per employee is the new benchmark, and everyone tries to be as lean as possible in building a business. And we've reached a thousand people. And I truly feel that we could be doing much more with our team. I don't mean that individual people aren't giving their all. Our people work very hard. But when you reach this scale, to do more, you need to work much harder on alignment, prioritization, and a clear understanding of what the business should even be doing. And it used to be simple. You could just walk into a room and say, "We're doing this." And everyone heard it. But now, when you have thousands of people spread across a hundred different cities, including a field team, it's very difficult to keep everyone synchronized and on the same page. So, when we reached a thousand, my first feeling was this. We need to be very, very mindful of how we organize the team now. Because before, the organization of the company wasn't on my mind at all. I always thought about the product, the technology, beating competitors, attracting customers, not about how we would organize the company. And now it's super important, and honestly, we still have work to do on that. Do you like this part at all? Because on the one hand, it's every founder's dream, and on the other, a nightmare. You always think that one day you'll grow to a scale where you solve problems for so many customers that you hire a thousand people, and then you reach a thousand and think, "Oh my god, what have I created? It's like a battleship. You can't even move your own company." Yes. Yes. And yet, the things that got you here, those very founder skills, we know each other well. You're not the kind of person who's eager to implement a lot of processes everywhere. Really, that's not something you enjoy doing. Yes. Yes. It's interesting, actually. In fact, I'm more annoyed by the lack of processes, because the lack of processes often leads to us having to do much more work. I have to answer the same question 100 times because we haven't documented it and made it available to everyone so that people can do the work without checking with me every time. So processes are actually an important part. That is, you can't be like, "I'm a cool guy, I'm not about processes," and then, like, somehow everything will work out. That's not enough to build a company. A huge part of a leader's role is to be able to organize, to be able to build processes, to form your vision of how work is done within the company. But your question was a bit about something else. Do I like it? And the answer is no. I understand that it's necessary, but I find it difficult to truly excel at it and lead the team so that it all really works. So yes, that's true. I think as the company grows, a person must grow with it. For example, forcing yourself to learn and adapt. That's what I'm working on a lot right now. Sometimes with success, but more often through mistakes. But I think that's how you learn. Today, essentially, every startup, every company I talk to, wants to be like Glean. Now it's something like Glean 4X. And on the one hand, that should be amazing, and on the other, just crazy. Even I, considering how closely I've seen this journey with you, think: "Do you even understand how difficult it was to get here?" Because even now people say, "Wow, it must be incredible to work at Glean." And I think, "Well, yes, the company is successful, but the actual experience of working at Glean has always been that it's not enough. And for years, it's been exhausting." And I'm interested now, with you having a thousand people, people probably come with the thought, "I'm going to work at this cool, trendy company." And it seems to me there's a dissonance between the reality of what it means to work at a company like Glean. And it's exhausting. Do you feel that? Yes, of course. Startups in general are very hard work. At the same time, there's a lot of joy in it. You have people who are truly motivated and united by a common mission. And, for example, as an engineer, as a developer. You really like that feeling that we're all in the same race. We're building a great product, we'll win, we'll attract customers. There's something truly exciting about this path, but at the same time, it's like high-level sports. Being an athlete isn't easy, you have to work very, very hard. People get tired after 4-5 years, the work doesn't decrease. And for us, that has been absolutely true. Well, you know, are all people happy inside the company? Do all people constantly feel energized? It's difficult. You really have to force yourself to feel it. Working in the industry doesn't make the task any easier, because it's perhaps the most fiercely competitive industry there is. There's so much change and so much competition that you don't feel like you can get off the treadmill even to grab a drink of water. That's the feeling of working at the company. But at the same time, I think it all works because we know there's a great goal ahead, and we're doing something big and meaningful on this path, and that motivates us, it keeps us together. And the last thing I would say, in response to your question. I always tell our team every week or every other week that it's going to be a tough and, in some ways, long journey, and that tomorrow won't get easier. It will continue to feel the same. That's the essence of a startup. You're constantly in a mode of hard work, so you need to find something that makes today interesting and enjoyable for you, rather than living in anticipation of tomorrow, which you'll supposedly reach someday. How long has this mode been going on for you? 7 years for you? 6 and a half. And what can you even do about it? I mean, you can't just leave. You can't just move to another company. You can't stop developing. You can't not learn processes and follow them. You have to do all of this, because otherwise the company will go under. And you can't let that happen, right? I don't know. The rubric went public. You founded the rubric. You've already made enough money. You wake up every day and think, "Okay, let's go." You just work hard again and again. Day after day, year after year. It's been 6 and a half years. And there's no end in sight. Doesn't that exhaust you? It does. And that's where I return to what I was talking about. You have to learn to make today interesting for yourself. I'm not waiting for some tomorrow when Glean will have a wonderful result. We will definitely succeed, and suddenly everything will stabilize. The business will just run, and it will be easy for everyone. That doesn't happen with any business. No matter how big you are, you can go public, but the complexity of the business and the level of responsibility you take on will be exhausting. It will be an endless series of work, stress, and everything else. So you just have to figure out for yourself why you're doing it, what's important to you, what else you could be doing. For me, for example, I know if I don't have a task to solve, my life will be unhappy. I don't have hobbies like traveling the world. I would just get lost if I had nothing to build. I need to create something. I really need to build something. I think that's how it is. For me, it's fundamental. I need work. Every minute I'm awake, I need to be doing something. But for me, it's enjoyable. It's enjoyable because, first of all, I'm mission-oriented. I like what we're building. I know we're creating value. I know we're changing how people work, making them more productive, and making their work more enjoyable, helping them achieve results. I also really enjoy just spending time with my team. Sometimes, you know, I won't go into detail, but some conversations completely drain me, while others, on the contrary, energize me greatly. And I try to plan my day so that there's a little bit of both. I need a dose of both. And don't you feel that you don't have a group of co-founders now who are really helping you? You have a management team. The company started with people, but ultimately it's your company. There's no getting away from that. That burden weighs on you. Like, the company is valued at $8 billion or so at its last valuation. And it doesn't look like growth will slow down anytime soon. Glean continues to accelerate its growth, and you're under pressure from all sides. Open AI wants to be Glean. Flexi wants to be Glean. Everyone wants to try to build a product like Glean. And you face very tough decisions, and have for a long time. At the same time, there are no precedents for being a company like Glean. I'm not even sure that even if there were more people in the company, anyone could come up with these decisions. They just don't have the right answer. Does that pressure you or not so much? Well, managing a company is difficult. I don't know how much more difficult managing Glean is than managing other companies. Most likely, it's about the same. Every company has its problems. We, for example, have a lot of competition now, but there was a period when we had no competitors for many years in a row. Then our problem was that we literally had to evangelize the market and convince people that a product like Glean would really help them and their employees. Now you don't need to convince them that such a product is needed. Not only us, but also 15 other large software companies are educating the market that an investment in a tool like Glean is necessary. Therefore, demand is huge. But now we have to compete and convince customers of something else, that we are better. And we can do that because we have more experience than anyone else in this product and in this field. We've been working on this problem the longest, so in a sense, opportunities for us continue to grow, challenges also continue to grow. And it's tough, it's difficult, but I have a great team. There are people who have been with the company since day one. My co-founders, a key part of the founding team, our leaders. And we are all here to succeed. I think we will succeed because we are focused. We are more focused on search and enterprise and platform than anyone else. And therefore, we feel quite confident in our chances of continuing to develop successfully. How crazy is it that everything is constantly changing? You're right. For years, you were the only players in the market. And there was a clear goal. Every day you got a little closer to that goal. Large language models in their current form simply didn't exist then. It was about connecting all applications and providing transparency across the enterprise. It was easy to explain, easy to understand. The applications were more or less the same. Employees adopted the products, they liked them. No one else was trying to do what we were doing. And now Open AI is trying to do it. New models are appearing, people are starting to build agent solutions on top of Glean, and now you're competing with a hundred different vertical agents. This pace of change is just insane. You are indeed one of the leading companies. And the price of that, the speed of change is higher than ever before. Yes, all teams in the company are under immense pressure, take our R&D team alone. There are constant problems. For example, when you think about creating a new feature, a new component. The question arises, is it worth doing at all, or will large language models soon become smart enough to do it all themselves? And we are now living in an AI world where technology is moving so fast that you can't even think in terms of: "We're building a tech moat." You're just building things in a certain way that over time gives you power because everything is changing too fast. In fact, I think any moat you think about becomes a burden in some way, because you need to evolve, move fast, change the product, change the technology to adapt to the new base form you talked about. It's difficult, yes, but we have a good team. Our engineers are capable of working in an environment with a constantly changing foundation. And so far, so good. People really like the Glean product. I was just on a call with one of the largest VC funds in the world right before this conversation, and they said the same thing. For them, search and platform are a priority. They've worked with large players, and they haven't worked out. And we are already implemented in many of their portfolio companies, and they like it. So they are happy to work with us. And moments like these also happen. They bring motivation back to the whole team and remind us, let's keep winning. >> Subscribe right now to my Telegram channel via the link in the description. I've prepared for you the top three materials that, in my opinion, everyone should know. First, a map of a hundred top AI startups – this is the future in one picture. Second, a forecast from an insider at OpenAI, who, even before the appearance of ChatGPT, predicted everything that is now happening with neural networks. And this year, he released a new forecast until 2027. And third, the most powerful – this is my analysis of an essay by the founder of Anthropic, who is essentially the second person in the world of artificial intelligence. He laid out step-by-step what will happen in the world in the next 5 years, and most importantly, what the universal AI, which everyone fears or awaits, will be like. Go to the link in the description. Your comment about moats. The moment you think you have a moat, you're probably already in trouble. This is because you're generally not sure if a moat is even possible when large language models are improving at such a breakneck speed. Yes, that's part of the reason. The technology stack is evolving at a pace we've never seen before. By default, I have this mindset. If you built something last year, it's already obsolete today. There must be a new, better way to do the same thing today. And if not, it's simply a lack of imagination. And that's why I constantly tell our team: "Encourage the creation of new moats just as much as throwing away parts of the codebase." Because if you don't throw anything away, don't get rid of the old, it means you are slowly but surely turning into legacy software. So this is a very important point, and therefore the idea of thinking differently is not about the code you've already written. What is it about then? Well, I think sometimes it's about how quickly you can delete and replace code. Yes, one of those is flexibility. The new currency now. And perhaps it always has been. But now it's more important than ever. It's the speed of adaptation. How quickly can you change, change the product, change the code. This gives you leverage to use the latest technologies that emerge. And another moat is your relationships, your relationships with customers. We work with the largest corporations in the world and are a key partner for them in their AI transformation. And we communicate this very clearly to them. Simply bringing you great technology is not enough to get a real effect from AI within a company. You need a partner, you need to work with someone, and we will be that partner. We will go with you on this journey for the long haul. We will bring expertise, we will bring general experience and knowledge gained from other corporate clients. But we will also spend time and constantly work with you. And it is these working relationships, how we approach clients, how we interact with them, and how we build roadmaps together to create value in their businesses. That's what they love us for. And if we expand on this idea, that code written a year ago is likely already obsolete for most SaaS companies today. Are you one of those who believe that almost all traditional legacy players, everyone who appeared more than 10 years ago, are now in serious danger because they have a huge customer base? The classic innovator's dilemma. They essentially need to rewrite their entire stack because it can be done much better. But to do that, they need to reassemble their entire existing business. Is that how you think about it? Well, I think there are two points here. First, people say that there won't be application interfaces, there will be a database with all the data somewhere in the backend, and on top of that, modern conversational interfaces, and all work will be done through them. I don't believe that. I think there will be many different product interfaces, many different products, and they will continue to exist. Corporate clients in the future will likely have even more products, not fewer. So I don't think that if you're a SaaS product, you suddenly become useless or unnecessary. But that doesn't mean you don't have problems. If you're a SaaS company, you can't remain static. You need to really go and innovate, update product capabilities, embed many native AI functions that meet new user expectations. But I believe that we will continue to have amazing vertical solutions for various business tasks that SaaS companies will offer. They just need to change, they need to use all the powerful capabilities in new product scenarios. And I'm not sure that this necessarily means rebuilding the existing business. I think it can be a great addition to what you can offer your clients today, considering what you see. And given that you are perhaps one of the technologists I admire most, do you think this music will stop soon? Like, will we continue to see things heat up? It's really hot in the Valley right now. Do you think it will get even hotter? Based on what you see in technology. Yes, well, if we're talking about an AI bubble or not, about valuations. I'm not an expert. I honestly can't judge that. I'm not even talking about valuations. The models continue to improve at the same speed. Yes, I think models will continue to improve. Moreover, it amazes me that the same basic technology scales. We've made it work on a larger and larger volume of data, on larger and larger systems, and until some fundamentally new alternative technology emerges that offers a different way to build these models, but it will appear. It's the human spirit of innovation. We will see much more. So I have no doubt. Opportunities will continue to grow more and more in the coming years. But an even more important idea is how much we are actually using what AI is already capable of today. I would say we are not using even 1% of the current capabilities of these models. Not even 1%. Because now models are designed in such a way that you still have to do a considerable amount of work on top of them to truly create products or solve problems that bring business value. Therefore, I can imagine a situation where there are no innovations in AI models at all. Let's say they stay at the level they are today. Even in that case, in the next 5 years, we will see huge growth in products across all industrial verticals. And from the 1% of current capabilities that we are using now, we will move to 10%, to 20%. So that's worth remembering. I think we sometimes get too hung up on the question of whether the improvement of basic models is slowing down or not. For most business tasks today, it's largely not that important. Do you have any idea, a framework, of how you Because in some cases, model providers are both your best friends and your competitors. And, frankly, I preferred the world when there were no competitors, not when you compete with players like that. This isn't some hypothetical Josh Mo as a competitor. Yes, these are the best companies in the world entering this field and trying to create Glean. But, you know, yes, that's one of the things they want to do. Obviously, all these companies are much larger than us. They have amazing capabilities, and for them, this has also become one of the important directions. So it's quite delicate. Yes, yes, I would say there are two points here. First, we are very close partners with all these model companies. We collaborate with them on a technical level. For example, we tell them in which areas models don't work very well yet. And we have a really good interaction in this regard. Of course, we are also their clients. We generate a large volume of usage and a large volume of revenue for them. And this concept of competition has become widespread because, essentially, every company today is expanding its product coverage. Everyone is trying to do more because AI really allows you to do more. Because of this, the number of overlaps is also growing. But I think this is largely a temporary phenomenon, and eventually everyone will find their own lanes to swim in again. And, for example, if you look at how our company is built, I believe we should be fully complementary to all model providers. We solve a different type of problem. And there was a time when we also trained models, and we still train some small ones. But my feeling, my hypothesis is that we will stop doing some of this work, and model companies will stop doing some other part. Why? Because ultimately, you realize, you can't do everything. You need to focus, you need to go deep to compete. You need to become really, really good at something specific. And if you try to be everything to everyone, you simply can't compete with someone who is focused on a narrower task and delves deeply into it. It must be a very interesting moment in time for you to observe what OpenAI is doing. They are incredibly ambitious as a company, and perhaps, as you said, it may not pay off. Over time, they may narrow their ambitions, but I wonder, are we doing enough when they are so ambitious? It's a strange feeling because you're like, "Wait, wait, we have to keep doing our thing as best we can." But damn, their ambitions are impressive. I didn't even mean that they would reduce their ambitions or narrow their focus. I rather think that over time, like any company, you make many bets, and then you start to consolidate them and delve into the areas that are most important to you. That, in fact, was my comment. I think that's what will happen. As for the question of whether we are doing enough, you know, internally, we listen to our employees, and they actually have the opposite complaint, that we are trying to do too much. We are a horizontal platform. We integrate with all different systems. We can come in and build agents for HR teams, for IT teams, for sales, for customer support. And therefore, it seems to me, our problem is the opposite. Aren't we trying to bite off too big a piece? And we are dealing with that now. One of our strategies is to be precisely a horizontal player and build very deep partnerships with all other vertical product companies. And, for example, how we, for example,
Our goal is to make it so that if you use a CRM product, if you use some kind of engineering system, within these systems we truly add value. We want to build a platform that provides both the deepest and broadest context of corporate data within each individual system. That's how we're thinking about our strategy right now. And in this way, we don't need to solve all the problems. We don't need to build the best products for every function, for every department, for every vertical. Instead, we become a platform that enables these user scenarios. We do a little work, add value, but maintain a focus on being precisely a platform. Have all these changes affected how you think about talent? Let's take, for example, engineering talent, considering Cursor and others. Has this rethought how you evaluate and perceive talent at Glean? One of the things we've started doing now is we test literacy levels for all roles. And the goal of the literacy test is not to understand if you are an expert in [AI/LLMs] and we don't expect that from anyone. We are trying to see if you are curious. Are you interested in what is happening right now? After all, this revolution is unfolding before our eyes. How deeply have you immersed yourself in it? How much has it awakened your curiosity? We want to see what you have already done with [AI/LLMs] and how you have used it. And this is part of the changes in our interview process. We now pay attention to this. Will a person be AI-oriented or not? Will they have that mindset? I want to do things in new ways, not the way I did them before. This is perhaps the main change that I am trying to implement in almost every team. But if we talk about the foundation, how we hire people, it generally remains the same. We, like everyone else, look at the standard things: hard work, talent, teamwork, and all that. But we always had one more skill that we tested and that was very important to us. It's the level of passion for the product we are building. It was important for us to understand how much a person shares the company's mission and is ready to stay with us, going through ups and downs, without naming names in general. I'm interested in senior engineers at Glean, those who, even before the advent of large language models, were obviously the best engineers, say, with twenty years of experience. Have you noticed any changes in how such people are valued in the organization compared to those who essentially grew up with Claude and Cursor? What do you think about it now? Before, everything was quite straightforward. It's interesting how this has changed for you. Yes. The use of [AI/LLMs] and my co-founders Tony and Vish. For them, the application of [AI/LLMs] is a kind of benchmark. Their personal assessment now is this. If a person doesn't use these tools, it indicates a rather narrow mindset, and therefore they highly value it when everyone uses [AI/LLMs] tools more and more. If we look at the patterns we see, especially among younger employees, people who are just coming in and for whom Glean is their first job, they become active AI users because, strangely enough, they are lucky. They don't yet know how to do their job the old way. They don't know traditional approaches and therefore immediately rely on AI to help them. And this is actually great. The younger generation automatically becomes AI-first. And then, if we talk about our team responsible for experience, we saw this. First, there was a growth curve in efficiency, and now we see, conditionally, two groups of users. One actively uses AI, the other is not so active. And we are trying to understand, because, honestly, we don't see any huge difference in productivity between these two groups. There are really great engineers who haven't integrated AI into their work as much, but at the same time, they still remain productive because, if you remember, AI has become very useful in writing new code. But the bulk of the work for the most senior engineers is not writing new lines of code; it's debugging, diagnostics, troubleshooting, it's thinking and designing components. And therefore, even without actively using code generation tools, they still remain our best specialists. Arvind, earlier you talked about processes and about writing, about how important writing is to you. We've been in many interviews together where candidates ask me: "What is Arvind like? How does he work?" I always tell them: "Before you go talk to him, put everything down in writing so he can gather his thoughts, understand, and see how you think." Just talking is not enough to understand your thought process. He really needs to see it written down, and because that way it's easier for him to understand your thinking process, and because it helps him to process his own thoughts and come back with clear questions." That's a fair characterization, actually. I'm surprised you understand it so deeply. That's what I need. But we've been doing this together for 6 years. So, yes, probably that's true. I think there are two parts here. The first is my own limitations. I need to read. I might have ADHD or something like that, but I can't process large amounts of information by listening. I need to read, which is why I prefer people to write for me. But it's not just important for me. I believe that writing is the best way to truly initiate deep thinking. And this applies to everyone. I don't think this is exclusively my peculiarity. I realized this by observing some of the best people in the industry. Back when I was at Google, I heard people talk about how, in order to communicate effectively and make strategic decisions, if you don't pick up a pen and paper and write down your thoughts, you will definitely miss something. You will be scattered, you won't have a coherent and cohesive way to convey your idea. I believe this is a skill that everyone should develop. And here arises the question: "What is the role of AI? If AI is now writing most of the texts, will this skill disappear or not?" I have thoughts on this, but in any case, you are right. For me, it's really important to see things in writing. What are your thoughts on this? AI? Yes, I think AI can provide a good start and help overcome that very first barrier. Whether you call it writer's block or simply a lack of energy or time to gather your thoughts. So AI can be a very useful partner. For example, I do it myself now. I just throw some raw, unstructured thoughts into AI and say: "Listen, put this into some structure. I get some artifact, and then I start working with it and deeply invest my own thoughts into it." Therefore, I believe there is a world where people will learn to use this technology correctly. Not in a way that creates what is now called "work dumps." When you generate huge amounts of information with AI, but you yourself are not even its reader, but just dump it on others to figure out. This, in my opinion, is one of the bad trends that are currently emerging from AI. And something needs to be done about it. One more observation. About your style, as it seems to me, tell me if I'm wrong, you have a real allergy to people who have achieved something and use their past achievements as a right to automatic reputation and trust. I don't know how best to describe it, but you have such a characteristic. You rely very little on what a person has done before. Moreover, sometimes it's even perceived as a minus, because you have the question: if you were, if you were so successful, why would you go to Glean and work hard here, if hard work is a basic value for us? Maybe I'm wrong, tell me. I wouldn't say that a person's achievements automatically make me think badly of them. But the truth is that for me, hunger, an internal drive, is a very important factor that helps people do their best work. And with very successful people, you need to be careful. I have a feeling that they might be less tolerant of what they have to go through in a company, because, essentially, they don't have to. They will more often ask themselves: "Do I really need all this? Do I need to work this hard?" And, of course, this is something I think about. One more point. Today, much of this experience, if it becomes the right way of working for you, starts to be perceived as if everything should be done this way, because you've done it once, not twice. And it has always worked for you in the past. But for a company like ours, this may no longer be the most appropriate approach, because the world is changing, organizations are changing, how each individual function is structured also needs to change. All these proportions, conditionally, do you need one solutions development engineer per one account manager, and so on. All these traditional metrics and processes, in a sense, are no longer directly applicable. You need to have a mindset that allows you to do things differently. Because much of the work that previously required people can now be done with AI. Therefore, it is very important to have an open mind. But at the same time, to be completely honest, there is no substitute for the fact that a person who has already achieved success and seen a lot possesses a value that others do not have. This experience is invaluable. I really appreciate it. It's just important that this experience doesn't become a ceiling for the person. That is, the person must still be able to reinvent themselves. How do you use AI personally at home? How do you apply it? Do you have any favorite use cases? Today, I've been asked this question several times, and it's made me realize that outside of work, I use AI very little. You're running around, choosing which model to run on the backend for Glean, and simultaneously making new requests. Yes, that's about it. How much do you work in general now? I think I work almost all the time I'm awake. I take breaks, but a break for me now is precisely the moment when I'm not trying to use AI even more, but watching TV, relaxing, and going to sleep. So my primary AI experience is work. I feel that it is precisely this year, perhaps in the last three to four months, that my own habits have fundamentally changed. Before, I used AI, of course, I used Glean a lot, I used ChatGPT mainly for knowledge retrieval. I have questions, I need answers. I'm trying to find someone to talk to on a specific topic. These were such questions, but lately, I've moved to the point where Glean has become a more powerful colleague for me than any other colleague. I don't want to diminish our excellent team with this. But there's something about AI that makes it a very effective personal colleague or companion. Now I have a lot of complex work. I constantly feel guilty about the list of questions that continuously appear in my head. My brain never rests, it's always working, and it bothers me. Questions come in an endless stream, and I feel awkward sending them to my entire team because it takes time, people get distracted. And I realized that AI is actually incredibly good for this. I can ask deep strategic questions, and I can ask Glean to work on it. And given that it has all the context of our company and how our business is structured, it's just incredible what it can do. So this is my new work model. Let's say I have a project, I need to make some strategic decision, or I want to deeply understand how things are going for us, how the business is doing, at the most fundamental level. I first ask Glean to conduct in-depth research and provide me with a report. A two-page report that will contain enough information for me to form my own point of view. And only after that do I start interacting with the rest of the team. Thus, I come much more prepared, and I am much more precise in how I use or take up people's time. And moreover, in many cases, I actually share the results of the work with the team, to also change their habits, so that they don't start work without AI and artifacts, and thus shape this behavior. And this is actually incredible, how much my work habits as CEO have changed. And, by the way, you remove a lot of bias. If you ask someone from finance or marketing about finance or marketing issues, they will think through the lens of their own worldview and then project it onto their manager. But this way, at least you have a relatively objective opinion, precisely comprehensive and unbiased. These are truly new, fundamentally new opportunities that AI brings. Have the questions in your head become louder, have they become more or less compared to 6 years ago? Are there more doubts, anxieties, paranoia, questions that don't stop your head from working, is it getting worse or better? For us, as a business, I think at first we had a lot of anxiety because we weren't successful, everything felt like an exhausting journey. There were many rejections, and therefore I had many questions. But then we started to see success, and it became obvious that we were the only player in the market, the best product, and all that. And now we are again at a point where there is a lot of noise in the market. There is a real existential risk for us. And this risk is that if you get distracted for even a moment, someone will overtake you, you have to be fully and constantly on guard. So yes, there are more questions in my head now than last year. And perhaps more stress. But at the same time, I also see that the opportunity for us now is 10 times greater than it was last year. Well, I really appreciate you doing this. For me, this has perhaps been one of the most valuable and rewarding experiences of my entire career at Kleiner Perkins. To be so close to what's happening with you. Glean, I'll tell you honestly, when I came to that Glean user conference, it was precisely a Glean user conference, and it was completely packed with people, customers. I couldn't believe it. I think I pulled you aside then and said something like, "What's going on? I just can't believe it." It's amazing. It's been a great journey and a great conversation. Thank you, Arvind. Good to see you. Are you hiring now? Are there any open positions? Yes, we are hiring across the board. Hiring engineers, a lot of people in sales, also leaders, quite a lot. For example, we are looking for a leader who will be responsible for our federal sector. So yes, all vacancies are posted on our website, and the company will definitely continue to grow. Most likely, we will double again this year. Wait until the number 2,000 appears before you. That will be truly frightening. Thank you, my friend. I really appreciate it.