AI Operating System: Building a Company OS with Claude and Agents

You've got these people who are these 1% AI users. They're highly AI-pilled. And then you have the 90% to 99% of the rest of the organization who isn't sure what to use when. Meet Jay-Z. She's the chief product officer at Loro, the $100 million AI time platform. She teaches PM at Stanford. Before Jay-Z, she was the CPO at Linktree, and she's led product at Airbnb, Webfloat, Dropbox, and WeWork. You do something pretty crazy in your interviews. Can you tell me how you interview people and really find the gem, AI-pilled, super senior ICPM? The fundamentals and the principles have never changed. In fact, they're even more important than ever before. But the tools and the way you operate, that's radically changed. How should people be thinking about, in an AI-native organization, this is the role of a PM? And so those are the four levels. Level one is you're talking to ChatGPT. You're talking to Claude. You're really using AI kind of in chat mode. Level two is where you start to automate a workflow. Level three is when you start building apps. And then level four, I would say, is where you're actually building, I call it, shared apps. How does someone start from step one? What is the process somebody needs to go through in order to build up and create their own company operating system? Before we go any further, do me a favor and check that you are subscribed on YouTube and following on Apple and Spotify podcasts. And if you want to get access to amazing AI tools, check out my bundle, where if you become an annual subscriber to my newsletter, you get a full year free of the paid plans of Mobbin, Arise, RelayApp, Dovetail, Linear, MagicPatterns, DeepSky, ReforgeBuild, Descript, and Speechify. So be sure to check that out at bundle.akashji.com. And now into today's episode. Jay-Z, I've been teaching people a lot about how to use Cloud Code with personal operating systems, with team operating systems. You guys at Laurel have taken it to a level I have not seen before. You guys have built out a company operating system. Can you show me what this is and what it does? Of course. All right, let me screen share here. Okay, let's start here. Let's go to GitHub, our favorite place. And so you'll see here that we have in GitHub a company-wide operating system where for every single function in a company, customer success, data science, design, engineering, finance, imitation, legal, marketing, we have essentially all these folders that share how do you think about each phase of work that that function does. So in customer success, you do account management. And within account management, you're thinking about renewals, upsells. You do a customer enablement. And within that, we essentially work with our customers. We do office hours. We help them with rollout. We do training and onboarding. Each of these folders have a skill, and I think for those of you who are less familiar with GitHub, we'll actually hop over here to something that is very familiar, which is essentially your file structure, your folder structure. And so going to customer success, you can see that each of these folders have a series of folders that are the activities that they do. And then within each of them, they have skills. So how do you actually think about creating the right assets for the negotiation support or the right references? I'll go back one more. For renewals, right? What is the skill file there to really think about how do you walk through a renewal correctly with a customer? And now you're like, okay, cool. You have some folders in GitHub. You have some stuff that you can download. How does this all come to drive real change? And the way I'll talk about this is at the end of the day, we all live in some form of email or Slack. And so what I'll do really quickly is I'll open up my Slack. And again, this is not real data in the sense that we do have very sensitive data that I'm not gonna be sharing. So this is a little bit more mock, but it shows you exactly how our team operates. So for example, every single morning, every person on a lot of these customer-facing teams, right, they're highly repeatable motions. The more we can sing from one voice and say the same thing, the way we can create consistency and the awesomeness of the customer experience, that makes your company much more unified and it's a big part of the brand. And so when you think about that and you think about a customer success person waking up in their day and really seeing, let me go here. This is an example for customer success. Here's your calendar. Here are all the meetings that you have, the check-ins that you have, the onboarding sessions you have. This is something that a lot of people are building. This example of a chief of staff light concept, but what we're now doing is we're integrating all the skills. So for example, when we do a handoff, when we do a session prep, all of these are actual skills. And what happens is then when anyone is using Cloud, for example, I'll just go into, I'll go really quickly into the organization settings and I go into your skills. You can start to see that you can upload all of these skills into your company context. And as a result, when you're going through your day, you can essentially say, great, I'm going through my day, I'm doing all of these things. I will use these skills so that I no longer have to spend all the time creating that one deck or spend all that time creating an email. It is actually something, you know exactly what skill to use when. And I think that's the biggest thing that companies struggle with, which is you got these people who are these 1% AI users, they're tinkering with their workflows, they're highly AI-pilled, and then you have the 90 to 99% of the rest of the organization who isn't sure what to use when. And so as a result, you can actually integrate your skills, again, at a company level. So across every single one of these functions, going back to files, each one of these functions and all the activity that they do in order to be able to understand what skill should I be using when and where should I be spending my time? Maybe the last thing I'll just show to really bring this to life is every single company, you can map every single function's work to what I call an ontology. So in sales, all of the work in sales maps to these categories that they're supposed to be doing. And within each category, there are a series of tasks that happen. And this is actually what has informed the ontology that I just showed you. We've done the really hard work of mapping out, okay, for every single function, again, I'll scroll through this, marketing, sales, customer success, implementation, design, engineering, so on and so forth. These are the things that we believe that each function should be doing. How do we actually create a set of skills for you to do the things that we want you to be doing more and to also automate the things that we don't want you to be doing anymore? So I'll go to product, which is a lot of the audience here. In product, what's really interesting is that you should be spending your time like an engineer in many ways. And we talk about this later where the ontology or the work map of a product manager is starting to look a lot more like an engineer. But there are a lot of things that used to be in the day-to-day of a product manager, doing competitive market analysis, doing these, all these like writing for stakeholder management or really mundane, tedious organization, getting people on a phone, synthesizing feedback, et cetera. All of these things, as we all know, are starting to get automated. But again, it's automated in a really lumpy way where one PM might be doing it really, really well and another PM might not be doing it as well. So what we can do here is when you onboard everyone with a company OS, again, going back to this GitHub and going to, let's say, product, right? You can start to say, hey, these are all the playbooks, all the skills that I wanna give every single person on my team. And then when they come in for their daily briefing, what ends up happening is that they are able to see their day at a glance and we essentially tell you where you can automate your day. So you take the thing that is essentially designed by the 1% of any given function, the person who is playing around the most, and you're able to spread those learnings throughout the entire rest of the organization. Wow. I think this is so powerful because we all have been working in different teams where there's that one person who's got their skills locked, but if they're just compounding in a bucket, then nobody can really benefit. This company OS, this is bringing that power to everybody. Now, you guys are an AI native company. You guys are an AI company yourselves. And so you guys would have certain advantages in building this. How does someone start from step one? What is the process somebody needs to go through in order to build up and create their own company operating system? I like to think about it as three different steps. And so let me screen share again, and I will share how do I think about essentially getting your steps in, going from most simple to most advanced. So the first way to think about this is how do you just start small? What is one workflow that you or your team does that is incredibly tedious that you shouldn't be doing again? So typically for many, many functions, it is I write this email and I want this email to have a template that is automatically kicked off for me when X, Y, Z things happen, or there's a sequence of things that happen. I don't wanna input my data into a CRM anymore. I want that to be automated. So there's some degree of thinking about what is super mundane, takes a lot of time out of your day to day. And if that were to be automated away, you'd be thrilled about. And I'll give you one very product oriented example, which is there are so many companies out there, so many PMs out there that spend a lot of their days responding to questions, escalations. So the sales team comes into a channel. I'm notoriously bad at my inboxes. I guess there's a version of that where I seem cool and unavailable, but the reality is I miss sponsor emails, guest pitches, and stuff that my team actually needs me for. So I got an AI assistant, the sponsor of today's episode, Ariso. Ariso connects to my email, calendar, and Slack. Then I just chat with it over Slack and it helps me with everything. It builds workflows to respond to emails, resolve customer issues, prep me for meetings. It actually comes to my meetings, updates its own knowledge, and remembers context from past conversations. So every time I talk to it, it already knows what I'm working on. I used to pay for Granola and Lindy separately. Ariso replaced both. One tool does more and it lives right in Slack where I already work. Check it out at ariso.ai slash akash. That's A-R-I-S-O dot A-I slash A-A-K-A-S-H. Here's the dirty secret about prototyping. You spend two weeks building a prototype. You validate your assumptions. Engineering loves the direction. Then what happens? You throw the whole thing away. Bolt changes this completely. When you prototype in Bolt, you're not building throwaway mockup. You're building real front-end code that integrates with your existing design system. So when you hand it to engineering, they don't throw it away. They ship on top of what you've built. I use Bolt every single day. I host my land PM job cohort on it. And honestly, I'm up till 2 a.m. some days just vibing in the tool, having fun, and building. That's when you know a product is good. When you're using it past midnight, not because you need to, but because you want to. Check out Bolt at bolt.news.akash. That's B-O-L-T.N-E-W slash A-A-K-A-S-H. Link in the show notes. Today's podcast is brought to you by Pendo, the leading software experience management platform. McKinsey found that 78% of companies are using Gen AI, but just as many have reported no bottom-line improvements. So how do you know if your AI agents are actually working? Are they giving users the wrong answers, creating more work instead of less, improving retention, or hurting it? When your software data and AI data are disconnected, you can't answer these questions. But when you bring all your usage data together in one place, you can see what users do before, during, and after they use AI. Showing you when agents work, how they help you grow, and when to prioritize on your roadmap. Pendo Agent Analytics is the only solution built to do this for product teams. Start measuring your AI's performance with Agent Analytics at pendo.io slash akash. That's P-E-N-D-O dot I-O slash A-A-K-A-S-H. Okay, so let's move into Slack and see what this might look like. You know, a lot of companies, if you just go into any Ask product channel or any channel, you see so many success folks, support folks, sales folks, other teams hitting up that channel, asking people, hey, I have a question, I have a feature request. And so the very small workflow that we did, and I'll go all the way down, is we created a Slack automation that essentially said, look, when a feature request comes in, we typically spend a bunch of time going back and forth, asking about, how many times was this asked about? Send me the gong recording where I can watch what the customer's actually saying. What is the impact of this for your customer, which requires some degree of judgment from the person managing the account? What is actually going on here? Give me some more details. All of those things usually require back and forth. So again, if I go back to this system of how do you think about a place to start, what is something that you do over and over again that you could really easily automate? And that automation for us was as simple as, hey, let's just automate what we ask someone to fill in, and then what often happens is you then have to triage it. You say, hey, is it for this team or that team? Is it for this PM or that PM? And what's the SLA to getting back to the requester on what we're doing about this feature request? And so all of that you can build into something as simple as Slack. So again, a lot of people have Slack, Teams, whatever it is you're using to chat with your teams. You can do something very simple where you essentially say, okay, great, I come in here. I'm going to automatically ask for all of this information. So what is it? Who is it coming from? What's going on here? It automatically assigns it to the person that makes the most sense to go look at this. And then it automatically creates some kind of ticket so that we can track it. And so all of that, again, this is one-on-one, I would say, is just a very small step in creating your operating system. So I start there. The next step is this idea of how do you start to really automate based on a bunch of things that your team is doing? And so the example here I have is, again, a team that usually has a lot of people, a lot of humans. At Laurel, we have a large GTM team, and within GTM, go-to-market, we have really awesome success folks who are essentially what I call time consultants. They're getting forward deployed into these organizations, helping them use Laurel as a product. And …[middle truncated — full text in the offloaded file]… Were you customer centric enough? Were you problem space first and not solution first enough? I just find it both so gratifying personally, but also such a great reminder of what product really is. I'll say one last thing, which is, what's funny is that I teach AI leadership through Reforge and that curriculum changes literally by the month. We teach it every six months and the amount of change between the six months is massive, but when you actually teach the fundamentals, when you teach what I call PM one-on-one, those core principles have not changed. You should still always never jump to the solution. Now that you can build faster than ever before, it doesn't mean you just build everything. What actually is important is to know why and for whom you're building for and what is it that you're trying to solve for and what success looks like and therefore, you actually know you've hit your target. What's really ironic is that through teaching all these different levels of product people over the years, I find that the fundamentals and the principles have never changed. In fact, they're even more important than ever before, but the tools and the way you operate and the way you can blast through the bureaucracy and feel empowered, that's radically changed. And so as a leader, the way you empower your team is very different. Do you have the right culture? Do you have the right team? Do you have the right space for people to even build? Do you have the right operating system? Do you have the right knowledge of what people are doing day to day? Do you have all of those pieces? That is changing dramatically, but in your actual one-on-one basics around what it is that a product person is supposed to be doing, the speed has changed dramatically, but what you're supposed to be doing at the heart of it, that has not changed. What a way to end it. All right, guys. We have hit a crazy milestone. We crossed 40,000 YouTube subscribers. We have also crossed 565,000 average views per listen per episode. When I started this podcast two years ago, I wouldn't have believed it. I want all 565,000 of you to flood Laurel's PM applications. For my money, this is like the coolest PM job you could possibly have. And I would say, if you are in a PM job where everything we were just talking about feels really foreign and like 10 steps away from what you are, find a job like this with an AI peeled CPO like Jay-Z. You are gonna learn so much more than if you get to this four years from now and then you learn it. Apply to Laurel, get her the best AI PMs in the world. Check out her class at Stanford if you are in the Bay Area, so you can really learn AI PM. And if you are a leader, check out her course and reforge. This is just me saying this. You can see how much value I got out of this episode. You can see I'll be writing about a company OS soon in my newsletter. Jay-Z has absolutely killed it. Thank you so much, Jay-Z. Thanks for having me. I hope you enjoyed that episode. If you could take a moment to double check that you have followed an Apple and Spotify podcast, subscribed on YouTube, left a rating or review on Apple or Spotify, and commented on YouTube, all these things will help the algorithm distribute the show to more and more people. As we distribute the show to more people, we can grow the show, improve the quality of the content and the production to get you better insights to stay ahead in your career. Finally, do check out my bundle at bundle.akashji.com to get access to nine AI products for an entire year for free. This includes Dovetail, Mobbin, Linear, Reforge, Build, Descript, and many other amazing tools that will help you as an AI product manager or builder succeed. I'll see you in the next episode.