#118 - Your Project Management Is Broken — Linear Fixes It

Show Notes

If your projects last longer than a few days, then you already know the problem: action items everywhere, people joining and leaving, updates getting lost, and nobody really having a clean overview of what is going on.

In this episode, Malcolm makes a very direct argument: the traditional role of the project manager — or Scrum Master in software teams — is becoming obsolete. Not because project management no longer matters, but because AI plus agent-native tooling can now do a huge part of it better, faster, and with far more consistency than humans can.

The center of this episode is Linear — the project management tool Malcolm believes is currently the strongest option for AI-native project execution.

Malcolm explains this through a real example: a complex EU-funded delivery made up of eight sub-projects, all running on a brutal deadline. In the past, that level of complexity would have triggered panic and a call to hire a dedicated project manager. Now, the work is coordinated through AI agents writing directly into Linear, while Malcolm can query the entire state of the project from his phone, generate Gantt charts, build dashboards, and even send updates while sitting in a car, walking outdoors, or preparing for a customer meeting.

That leads to the key concept of the episode: hypervisibility. Instead of project status being buried in weekly review meetings, PowerPoints, Excel sheets, or filtered reports, everyone — including leadership — can ask the system directly what is happening, what is blocked, who is late, what has no due date, and what the next steps are. That changes project management from a ritual of chasing updates into a live system of transparency.

The episode also lays out why Malcolm sees Linear as structurally different from older tools like Microsoft Project, Jira, and Asana. Those tools were not built for AI agents first. They can be made to work, sometimes painfully, but they are slower, heavier, more customized, and far harder for AI systems to reason across. Linear, by contrast, behaves more like an AI-native coordination layer.

And perhaps the most surprising part of the episode is this: Malcolm argues that using AI for project management does not make work colder or more mechanical. It actually gives him more space to be human — less mental clutter, less fear of forgetting something, more presence with family, more calm, more energy, and more room for better conversations with colleagues and customers.

---

🎙️ ABOUT THE HOST

Malcolm Werchota leads AI adoption programs for companies across Europe. After more than 15 years in international corporates and leadership roles, his focus today is practical AI implementation without the usual nonsense. He works with companies from manufacturing to pharma, from family-owned businesses to large global enterprises — always with a strong bias toward real-world adoption and business value.

---

🚀 RESOURCES FOR LEADERS

📚 Chief AI Academy — AI for Decision-Makers
https://www.werchota.ai/chief-ai-academy

👥 AI Leadership Community
https://chief.werchota.ai/getting-started

---

📬 CONTACT

LinkedIn: https://linkedin.com/in/malcolmwerchota
E-Mail: [e-mail]

---

🔎 TAGS

#AI #AICookbook #Linear #ProjectManagement #AIAgents #Hypervisibility #ClaudeCode #Codex #AIAdoption #EnterpriseAI #ScrumMaster #Leadership #Automation #FutureOfWork

What is The AI Cookbook Show by Malcolm Werchota?

Malcolm Werchota's AI Cookbook Show is where artificial intelligence meets authentic business transformation. Known for his direct style and willingness to show AI in action—even during live presentations—Malcolm helps organizations understand that AI isn't about replacing humans but amplifying their capabilities. From voice-note productivity hacks to real-time meeting intelligence, this podcast delivers actionable insights for immediate implementation.

2026_04_26- EN - E 118-Linear: Hey everybody, welcome to the AI Cookbook Show, the only bilingual AI show. I think it's the only one. In English, it's called the AI Cookbook Show and in German, it's called das KI Kochbuch. So today we're gonna talk about project management because I reckon anybody who listens to this is a business leader. And you know, you probably also run projects and I reckon your projects go longer than a day or two or three and maybe they go. for a week or two or maybe they go for months or years. And you know what the problem is. The problem is there are action items everywhere and new people come in the project, old people leave, et cetera. And nobody has an overview. And you could say, but Malcolm, we have project managers or we have scrum masters. And in software, these project managers are called scrum masters, right? And their entire job is... project management, like writing down action items, keeping people on track, etc. And today we're going to talk about the fact that I think... that this role is obsolete. I think that the role of the project manager is obsolete and that it can be done better by AI and AI agents. And I don't want to be too, you know, like, ⁓ they are AI agents. I want to be very, very, very concrete. So two days ago, that was on Thursday at 3 a.m., but it was then Friday at 3 a.m., you know, I'm working on a... and an EU-funded project, which is actually eight sub-projects with a very hard deadline. We only have two months to deliver this. And my new setup for doing such projects is on the left is written is code code and on the right is codex. And both of them are connected to the same subfolder. So they kind of working together on the project. But so that these different sub-agents can talk to each other and know what happens, and maybe one of them maxes out so I need to wait four or five hours on one of them. And so that I have an overview of what's happening, they need to write down the action item somewhere. And the way that they do it is they use linear. And I can tell you that five years ago, I would have had an absolute heart attack knowing that we need to deliver on such a project. My biggest stress would not be the delivery, like I think we can do that, but the biggest stress would have been like, how the hell do we keep ourselves updated on eight different projects on such a short deadline? And probably I would have said to the team, I think we need a project manager for this. But I can tell you that at the time, which was two days ago, but even today, I am super duper calm. Because... All this project management is done better than any human in the world can do it. It's been done by AIs. So every agent, every sub-agent, when they finish their work, they go and write down an update on Linear. And that update is faster than I could ever type it because different sub-agents work at the same time, which creates more context. And today on Saturday, well actually now it's Sunday morning, on Saturday we drove to Algoi. Okay, beautiful area, mountains, lakes, et cetera. And Masha says, hey, where do we stand with that project? And guys, no joke, you open Claude on your phone and you say, build me a Gantt chart of the action items. Where do we stand? Pick out the action items. Which ones are we late? How much time do we have? do the entire analysis and by the way, send Marsha an email. And with this, this proves to me that all of us here are not able to do project management if we don't use AI and AI agents. So in my company, we're working on more projects than ever. And every time we thought we hit the ceiling, I feel that we could take on double the work. And normally I should feel double the stress, but I'm calmer than I was in years. I'm Malcolm Verchota. This is the AI Codebook Show, episode 118. And by the end of this episode, you will learn four things. Why linear is absolute, excuse me for being so profane, the shed versus Microsoft project, Jira, Asana, et cetera. And why hyper visibility is a word that you need to introduce in your company. And finally why your scrum master is probably obsolete and why when you work like this it gives you more time to be a better human, a better father, a better colleague, a better friend and not less because you work more through AI. So yeah everybody, action let's go! Linear, linear, linear, linear. This guy keeps saying linear all the time. Okay, what is it? Okay, linear is a project management tool, a bit like a Notion, Asana, Jira, et cetera. The thing is, it's a startup which was founded in 2019 by a designer out of Airbnb ⁓ and engineers today from Uber, Coinbase, etc. And in 2026, it's one of the fastest growing pieces of enterprise software on earth. So 25,000 plus companies trust the stuff. They're making like a hundred million dollars of ARR. And their profits, listen to this, their profits versus last year are up by 300%. So it's not a small company around the corner, no. It's a company which today is valued at over $1.25 billion, which means it's already a unicorn. And that $100 million of ARR, that's absolutely crazy. It's done with a team of around 200 people. So who the hell uses Linear? Well, lot of companies. OpenAI, the company that does ChudGPT, they have over 3,000 people on Linear. Ramp, they're scaled from 5 employees to over 1,000 on Linear. But not only them, know, Versal, Coinbase, Cursor, etc. And roughly 50 % of all the Y-combinator companies use Linear. and nearly 70 % of all the top AI startups use Linear. So if you're working at these old school dinosaur companies, you're probably not using Linear. But let's look at three reasons, because remember, consultants speak in three, why Linear is so bloody cool. The first one is Linear is basically API first, which means I have never opened Linear on the web. Like I never went to Linear and opened and checked my action items over there. I have the app on my phone, but I rarely go inside. All my interactions with Linear are either through Cloud, Cloud Code or Codecs. Which means it's not an app that went and built an API. It's kind of like an API first. And because it's API first, it's like self-documenting, self-discovering. Basically, like that same API that linear people use internally, you also have. And which means the following. Because it's API first, it's great for AI and AI agents, which means I use AI to go and query linear, but also to write new action items or to say, go and check my emails every single day, pick them all out and write automatically action items that are in my emails in linear. And that just works. The second thing is it's like speed, speed and simplicity. In Jira, let's say you want to try to understand what the hell is even happening in my Jira project. How many issues do we have? How many this? How many that? You need somebody who's very good at handling Jira. Often you need a Scrum Master, right? Now, in Jira, you also need to fill out loads of custom fields that nobody wants to fill in. And Jira is kind of like a Swiss Army, a Swiss Army knife, you know the Swiss Army knife that you have, but it has like, I don't know, 17 rusty blades. And linear is... one single blade. And with that one blade you can cut Swiss cheese, can cut your salami, but you can also cut an entire steak with one single blade. And for those who didn't understand this, maybe the third one is it's really built for AI agents. How do we know this? Because in May 2025 Linea shipped an official MCP server. Who did they launch this with or who did they co-launch this with with Anthropic? So Claude, okay? And Linear and Anthropic kind of like got together and then they shook hands and they said, ⁓ let's make this work natively. You know really well, like plug and play. And it works so well. It's like not a hobbyist hack. Most old softwares will also have an MCP server. It's gonna be so slow that you prefer not working with the MCP server. And that tells you where this is going. This is not only, okay, let's go and patch AI into it. This is natively patching and working with AI. Yeah, not bad Malcolm, but I don't need this. We work with Microsoft Project. Bro, do you understand that Microsoft Project is nearly almost as old as I am? I was born in 1983 and Microsoft Project is born in 1984. And which means it runs the architecture of a Microsoft Project runs on something called the project server. And the API is a legacy rest. Which means by the time you get Claude or Claude code to go and speak to it, you could have rebuilt this entire project much faster in linear. So it's like a dinosaur software and it's not native to AI agents. And now really the clever people are gonna come here and say, Yo, Yo, Malcolm, we use JIRA, you know. Yes, know, JIRA is kind of cool and it was a cool tool kind of before AI agents existed. So now you're not anymore an AI Neanderthal, you're more like an AI middle-aged person, not middle-aged, like living in the middle ages, like in the 1400s. Was the 1400s the middle ages? I don't know. So anyway, so the REST API of Atlassian is super heavy because it returns a lot ⁓ of context. That is this Atlassian document format, right, ADF, which means It's multiple trip retrievers, verbose payloads, know, it's like nested JSON trees, et cetera. Which means it's super difficult for an AI to even read. Now, let me explain this. Now, when you take queries, and this was not one time query, it's like over 90 days they did that. When you ask a query to JIRA, it takes 3200 milliseconds to answer. That's not so bad. But Linear's GraphQL averages 47 milliseconds. Which means Linear is 70 times faster when you go and connect to it. 70. And the reason Atlassian will say that it's so slow is because every single Jira instance is hyper-customized, your story might be somebody else's tasks or somebody else's epic. And AI agents find it very, very hard to reason across this chaos. And I know that Atlassian shipped Rovo MCP server, so they're kind of catching up, but it's not the same. Now, Linea was built in 2019 by way by people who hated Jira. Like they wanted something fast. And at the time when they built it, AI agents did not even exist or they didn't know it would arrive. So it's kind of a bit by accident, serendipity, that they built a super simple software, which is API first, which is exactly what AI agents need. And that's why Claude or Anthropic went to partner with Linear because it just works. and the codeword is hypervisibility. Okay, now I'll tell you what I do in front of customers. So I will say, you know, like we would like to work with you and then I do the pitch and then I show two, three AI things. And then one of the things I say is, you know, I'm also very, very confident that we are able to deliver because I have very, very strong overview of what's happening in my company. And I take out my phone, Claude, and I kind of say, okay, ⁓ you know, like, Let's imagine how we will do project management for your company. Then I will say, look, pull up all my linear issues on Claude, on Claude, on the phone, you know, build me a Gantt chart, color red for everything that's 14 day plus late, orange seven to 14 days, and then yellow under seven days and green for anything that's still upcoming. And then I kind of say the next one, I'm like, ⁓ build me an HTML dashboard, you know, like. open issues per project, issues without an ISI, issues without a due date, etc. And then I say, ⁓ by the way, when we're at it, also please find the blocked action items. For each, try to infer why it's blocked. Because you have access to my emails and you have access to Linea and through the TADV API, you have access to all the calls. please try to find out have the customers already complained, where do we stand, where are we stuck? And then that thing works for two, three minutes. But when I prompt like this on my phone, on Claude, my customers know that they don't have that. They don't have a CEO in their company who can query the entire project management system like this while potentially sitting on the toilet. Do you understand this? You can sit on the toilet and do this entire analysis on the toilet. You don't need to open your emails, you don't need to open an Excel which is filled out by some project manager who works hours on it every single week. No. You basically open your phone and you talk to it. By the way, this works with Claude, works with OpenAI, with ChatGPT, it works with a local LLM that you could set up. But what are you doing instead, you and your team? Well, you're doing Friday review meetings. You know, where like 20 people are inside. Did you do this? Did you do that? Where do we stand? Did you send me an update? Did you fill out your presentation before, et cetera? And I'm saying, ask, linear instead. And that creates exactly that code word which is hyper-visibility. But what are you doing now instead in your company? So you're doing these big project management meetings. They are maybe on weekly basis. And 20 people attend them and then say, where do we stand? What do we do? You have people who read lists like you're paying people to read lists. Now, you must understand what you're doing there. because let's say the hourly rate of that person who's attending the meeting is 80 euro an hour. And 20 people are meeting. and it's a two hour meeting. So this is a 3,200 euro meeting. But if you say, that's not so bad. No, no, no. If they're doing this meeting, let's say whatever, 46 times or 48 times a year, let's say 48 times, then this is a 150,000 euro meeting. Like you're burning 150,000 thousand euro cash trying to do project management. And that is really, really, really wild. Because it is proven, there are many, studies on this, that 30 % of a project manager's daily tasks can be automated this way. Probably not full, but at least 30%. It's not only that, it's that when you use it with linear and AI and AI agents, your project status is live. Anybody in the company, the CEO of the company can say, what the hell is happening on this project, et cetera. And by the way, AI's are, you know, they're not so political because when people give feedback during such a meeting, it's very political. And compared to humans, I think they are much less unbiased. So by using AI, And AI agents, you skip this entire political theatre. Okay, so now I'm starting to get motivated. basically, AI agents are a thousand trillion better percent project managers than humans. We don't even need to discuss this. Why? Because humans are a pain in the ass. know, humans are moody, humans are difficult, humans have personal problems, issues at home with a spouse, with the kids, maybe somebody's sick in the family, but we bring the stuff to work. And then we complain and then we're unreliable. And then we go on vacation. Like we go two weeks on vacation. I know the Americans will listen to this, you guys don't, but like in Europe, we have whatever, five, six, seven weeks of vacation, but it doesn't matter. Even two weeks on vacation. So then we go two weeks on vacation. Who does the project management? And we forget things. My God, we're champions at forgetting things. So when somebody says, where do we stand? We take it personally. And if the CEO writes and says, ⁓ that project, I want the status update, I mean 20 people work on it. And then somebody works on the project update, cascades it one level higher and another level higher, two, three things are filtered out and then it's being sent to the CEO. AI agents don't. Which means Codex doesn't have a hangover after it drinks three beers with Malcolm. And Claude Cote does not fight with its spouse like Malcolm does sometimes. And these AI agents don't miss your daily stand-ups because their kids are maybe sick. And sub-agents, let's continue, they are not jealous when another sub-agent does the task a bit better. Hopefully you start to get the point that I'm trying to make. Structurally, the role of the project manager or the scrum master was created in an era when status update required humans. Because... You couldn't use software to pull data, summarize it, write a really nice report and ping it to the right people. But that era is over. Because now, I'm not saying you should fire all your project managers. I'm saying fire 20, 30, 40 % of your project managers. And then you restructure the role. And that PM is now not anymore a 20-year-old that came out of university. That PM is a senior business leader with gray hair who's at least 50 years old. What are you talking about, Look here. We see in companies that are very strong in using AI that that project manager role has bumped up two or three levels versus the level where it was held before. Before it was some young motivated person who came in and loves to spend a lot of time on Excel and loves chasing people up. And now it's a senior person. Because that senior person is not consolidating Excel sheets. No, Claude Code is doing that. And that person is not running seven meetings a week to ask where the hell are we. No, that person does a very, very big percentage with Claude and Linea. And then that person is used to give a strategic overview, to move people in the right direction based on deep analysis that they have done while querying code code about all the linear action items, probably while sitting on the toilet. Now, I don't want to be your typical consultant, a kind of him, but we know who always says, ⁓ everything is awesome. No, there are a few downsides, so let's go into them. By the way, there are two downsides for all the young people here out here who want to become consultants. When you go to your customer, you're always going to say three very good things, but less bad things. So we have three good reasons, but we have two things we need to think about. And by the way, they're not disadvantages, they're risks. By the way, you can turn any risk into an opportunity, but that's a separate discussion. So one risk or limitation is hallucinations because we have also seen it. So AI agents sometimes create wrong action items or they will invent a due date or assign it to the wrong person. So last week, Claude Code invented a, I mean, there was a linear issue, an action item, but it gave it to follow up with Helmut. A, we have no Helmut in this company. And b, even when I check my emails, funnily enough, we've never even dealt with a Helmut, even on the client side. So, the AI invented Helmut. Okay, but one thing that will save you a bit is when a human makes a mistake and you ask that human three months later, why did you make that mistake? Nobody knows. But linear has an audit trail, which means every time something changes, you are able to in six months, in one year, one and half years, go and check. On a PowerPoint you can check, on an Excel held by a Scrum Master you can check. But with Linear, you can't. And the other one is, you build a human reviewer, human in the loop review process. Which means I told you that Claude basically reads all my emails and then fetches out the action items every day at like two in the morning. and then hands this over automatically to linear. But it does not put it in the action items, it puts it in the backlog. So I can go through them and say, that's a good action item, that's not needed, that's not needed, or I haven't thought about that. So, funnily enough, three quarters of these action items that I picked up, I can mark as done automatically because I kind of know it's already done or it's somewhere else. But as a human, I would lack the discipline to do this. to go and read email and create action items of it. And that gives me the discipline. Because without those five minutes, me, like every other business leader, we're flying blind. And the second problem you will have is, like in the second brain that I kind of explained, know, some employees won't want to use it. Why? Because they're afraid of it, because it makes their work visible or their lack of work. And when you have that salesperson, the superstar who you don't wanna touch, who's like, I'm totally swamped. Linear plus code, code will kind of show you, ⁓ this guy or this girl has like three open action items since two months. Hey bro, you've been swamped with three action items, like, hey. And that visibility, this hyper visibility which we're speaking about, That is really a feature, probably in terms of ⁓ change management, adoption of technology, it's also going to be a blocker. Because these politically protected people, the people that have a godfather or godmother somewhere in the company, They profited until now from opacity. Remember that word that I'm not sure it exists, opacity? And now you have to decide what you want. You want to still work like you were working in the nineties in a highly fragmented organization with project managers who are managing your JIRA boards? Or do you want hyper visibility in your company, like a fluid organization? And I know what I want. I never, ever, ever want to hire a project manager in Wechota AI ever again. Okay, so what are you able to do or what am I able to do now that I work like this? Well, the first one is I have more energy. I can come home after work and be a better father because I used to come and then, well, actually I work from home so I just go out of my office. I still spend time with the kids, but maybe I'm on my phone, I'm thinking, ⁓ my God, I need to do this, I need to do that, need work, et cetera, right? And in that moment now, when I think I need to do this, quickly speak into Claude in one minute, I still need to do this, put this as an action item on linear, this is the start date, this is the end date, this is the customer, it's out of my head. And when I did the Scrum Master course, I'll tell you when I the Scrum Master course, it's been like 10 or 12 years ago. I did my first Grand Master course, they kind of explained stuff like that. Whenever you think of something, write it down right away, right? So, which means I can concentrate more on the kids. You know, I can concentrate more on my family. I think that I can put my kids better to bed and not be so stressed like I used to be. Because before I had this fear of missing something. Even or especially if I'm working on eight projects at the same time. Now I need to be less stressed because these AI and AI agents are constantly updating Linea. And if I close my laptop and I have two or three laptops and two virtual machines, even if I start on another virtual machine and there's a different account on Cloud Code and on Codex, I can just say like, go and check out Linea, see what we've done and let's continue working. And even if I'm in Vienna and then last week I was in Vienna for Big Bang that everybody knows, actually a couple of times and a buddy calls me in the evening and say, let's go out for a drink. Maybe before I would have said no, because of work or the stress of missing out or the stress of having to work. And last week I went without a bad conscience and then I have had a big headache, unfortunately. So what I'm saying, if I want to summarize it in a sentence is maybe something like technical leverage created by AI agents and project management. is creating space for more human depth and better human interaction. Because I really believe, and I hope that the CEO of Linea listens to this episode, I believe that Linea made me a better father, a better husband, and a better friend. And I'm not even selling you Linea. Now the CEO wanted to listen and now he's gone again. I'm telling you, whatever tool you pick, and we're talking only about project management today, must really be AI friendly and agent friendly. Okay, Malcolm, what do I do? Okay, three things. Remember, young consultants? So the first one is go and sign up for Linea. Like take one of your projects. You don't need to move all of your projects to Linea for now. Take one. And you know, AI features, they have this AI feature called like triage intelligence, et cetera, on the business tier, if you want like $16 a month and stuff. And if you want to go full AI value, go business. But $16 a month is. cheaper than a coffee at the airport in Zurich. You when you go to Starbucks and you drink a coffee and Zurich airport, I think it's like nine or $10 or something like that. And the second one is you open Cloud, you go on Connect Settings, Connectors, Linear, you connect to it. And the same is by the way in ChattGPT. Also possible in Co-Pilot Studio, a bit of a pain in the ass to set up, but it's possible. And then you start querying your project, but also entering the action items through the AIs. And if you currently are using whatever g-router so that exports your entire project as a CSV, give it to cloud and say replicate this on linear. And then you go on your phone and then you say these things that I told you, build me a Gantt chart, action items per project, me the blockers when my team is doing the least. You don't need to go to your Scrum Master or your project manager. You can do that. And I'm sure that you will get better answers from Claude and Linnea than your project manager is able to do it to you. And from that moment onwards, hopefully you will be converted by Friday. And if not, then I'm sorry, friend, you are a total neanderthal wanting to continue work in the old way of working. Now I'm going to stop, you know, know, ⁓ know, insulting everybody's listening to this podcast and bitching, et cetera. And maybe a way to do it is when a project manager leaves your company. Because that's an opportunity. Because the next person you hire, as I said, is not a project manager. It's that AI orchestrator. Half technical, half organizational. And I kind of explained why it has to be somebody who's senior or somebody with gray hair or no hair. But I believe the more senior this person is, this AI orchestrator on the project management side, the more you're going to get an input out of this. And if you really want to learn about this with your team, then come to our Chief AI Academy. The next cohort is starting in a week or two. And there we're going to explain to you how you do project management with Linear plus Cloud and Cloud Code. And also how you combine it with our second brain, the Wachotha second brain, to even get better context for that thing. So yeah, you see, now it's already 1.30 a.m. in Bregenz, but... Back to my time on Thursday evening when I was working on these eight projects at the same time with Claude and Claude Code and these sub-agents writing into Linear faster than I could ever type. I'm not stressed the moment I close my laptop because I'm not afraid where will we stand on Monday? I'm going to meet the customer on Monday because I know half an hour before I meet the customer I can say to Claude, build me an HTML dashboard for everything that we stand on this European Union project, action items, Gantt charts, blockers, next steps, et cetera, because everything comes from linear, because the agents, the AI agents, they logged everything. And for other people, maybe the code word today wasn't linear, but it was hyper-visibility. Because if you can say that in your company, everybody has hyper visibility on all the action items on project management, then you've made it. Then you don't need to do anything that I told you today. But I'm not saying everybody on your project. I'm saying even the CEO of the company, if he or she can query everything that's running in your project, then you've done it right. And I believe unfortunately that 99.9999 % of all the companies in the world, that is not the case. And if that is not the case, Then my friends, go and get yourself linear. So I'm Malcolm Verchota. This is the AI Cookbook episode 118. If you want to listen to it in German, this is the AI Cookbook episode 130, live from Bregenz. This is produced by Verchota AI, actually sponsored by Verchota AI, meaning my time and everything else that is needed to make this podcast work. So yeah, most wonderful greetings from Austria. All the best. Take care. Malcolm out.