- AI agents are rapidly transforming workflows beyond just coding, enabling small teams to manage large-scale operations and achieve significant productivity gains.
- The speaker highlights that AI's declining cost curve makes its widespread adoption crucial, fostering "tiny teams" with high revenue per employee.
- Crucially, agents free up human creativity and productivity by automating tedious "yak shaving" tasks and even managing complex data, shifting focus from code output to human output.
Agents for Everything Else — swyx
- The cost of AI continues to drop dramatically, roughly 100 times every 12-18 months, making its increased usage highly advantageous.
- "Tiny teams," defined as having more millions in revenue than the number of employees, can be built and scaled by leveraging AI agents for various tasks.
- AI agents are particularly effective at eliminating "yak shaving" – the tedious prerequisite tasks like installing dependencies or fixing environment issues.
- They can rapidly convert design mockups (e.g., Figma) into pixel-perfect, functioning websites, speeding up design-to-development cycles.
- Non-technical team members can effectively use agents by communicating in natural language, leading to increased productivity and even creative "fun" projects.
- AI agents can manage complex data problems, such as conference schedules, speaker changes, and sponsor coordination, acting as a flexible CMS replacement with code as the source of truth.
- Personal automation for knowledge work, like researching vendors or making purchases, can be offloaded to agents with web access.
- The rise of agents is shifting the focus from traditional UI dashboards to APIs and CLIs, as the primary user of services increasingly becomes other bots or agents.
- To adopt AI for replacing existing SAS tools, identify and systematically reduce the top three concerns associated with the change, while addressing employee concerns.
Tiny Teams — Teams that generate more millions in revenue than their total number of employees, emphasizing high efficiency.
Yak Shaving — The process of undertaking a series of seemingly trivial tasks that are prerequisites to a larger, more important task.
Coding Agents — AI tools capable of generating, executing, and debugging code, often to achieve a specified goal.
Pixel-perfect — A design or implementation that precisely matches the visual specifications, down to every pixel.
Red lines — Annotations or markings on a design showing precise measurements, spacing, and other details for developers.
ETL — Stands for Extract, Transform, Load; a process in data warehousing that involves extracting data from sources, transforming it into a usable format, and loading it into a destination.
CMS — Stands for Content Management System; a software application used to create and manage digital content.
Open Claw — A reference to agent architectures (like Open Assistant or similar frameworks) that aim to provide open and extensible AI agents.
Agent Experience (AX) — A concept referring to how AI agents interact with systems, data, and human users, analogous to User Experience (UX).
SAS — Stands for Software as a Service; a software licensing and delivery model in which software is licensed on a subscription basis and is centrally hosted.
The future of work has many paths. Our next presenter will discuss the path that he walked with Devon as he organized this very conference. Please join me in welcoming to the stage the co-founder of AI Engineer Conferences Swix. Hi everyone. I am not the chief air officer of the UK. Unfortunately he had to leave for personal reason. Thanks for staying so long. I hope everyone has a good time. Thank you. It's so enduring and heartwarming to hear from you guys. I'll take you a little bit into how we build AI Engineer with AI. It's probably the biggest revelation that I've had. So we've had a lot of really warm reception from you guys and I think it's really great. I think this is something that we really try to engineer and hopefully this is our first event in London. Hopefully you have us back next year. One thing I wanted for those who are newer to us, I do one of these key nodes every single AI. The first very first one three years ago, I talked about the productivity gain that you get from the increased usage of AI. The second one, we talked about how you should just use more AI because the cost curve of AI is going down roughly a hundred times per every 12 to 18 months and I think it's still continuing to trend that way. The third year we started to talk about tiny teams which is basically this definition that I had that teams with more millions in revenue than number of employees. I even curated an entire track at the world's fair about this where we sort of summarized it as the tiny teams' favorite if you're interested in building that. The reason I liked this emphasis is because I think people are maybe too egotistical about looking at, look for the one person billionaire or unicorn founder. Every company can have a tiny team whether you're smaller large. I think when I look at how we run AI Engineer me being the leadership of Ben Lee and myself, we are also a tiny team. This is us. It's just a nine full-time people and we are running a business that is more than $9 million so we are a tiny team. I wanted to show you the most significant changes in our workflow since we started this two years ago. By the way, this is our take in the AJIP Hill moment. Did you guys get the AJIP Hills? Yes. Very proud of this is my brainchild. If one of your co-workers is not sufficiently AJIP Hill, you should prescribe one of this. You are all AJI doctors now. Okay. Our stack was very stable and completely non-AI which is very ironic for an AI conference. We do Figma, React, SuperBase, Tido, Google Sheeds, Session and so on. I had this funny weird moment where I joined cognition and started using coding agents seriously at work mostly because they were free. I started adding it to the company's lack and then I started doing things that didn't show people, hey, here's how you use it to do coding on the company website. Oh, well, I'm good. Something strange starts happening. I started introducing this is a workflow of our contract designer in awful time showing me a Figma page and asking me to go through it and expecting that we would take a week, two weeks, four weeks to turn it into reality. I just added Dev into it and ultimately, before I had to add Dev into it, I had to cook up Dev into Figma and I'm not going to do that shit. So, core is doing it for me. You should use core for doing this. And which by the way, leads me to my first lesson, which is anytime there's random yaks shaving, I think one under-appreciated benefit of agents is that they save you the yaks' shapes, all the dependency tree crawling of like, oh, no, I have to do that first. No, I have to do that first. Particularly when it comes to installing dependencies of fixing pipe on dependencies, fantastic for that. And I think model productivity that doesn't sufficiently appreciate parallelism and not just autonomy, I think, and as a death of the yaks' shaving is not fully capturing the benefit of agents. So anyway, back to the agent story, hooked up Dev into Figma and in very short order, we have a perfectly functioning website that is pixel-perfect to the Figma. And to me, that was the surprise I've never done it before. You always mistrust marketing until you see it for yourself. And more importantly, our design knows very happy about it. And that's basically the website that you see left today when you go to AI.engineer. The other interesting thing that happened was that we started using it more, after one initial success, you started using it more. Something that you can't see because it's very small text, but I'm going to highlight for you is that that is 207 replies, just exploding in usage, like, what the hell? And when you dig into it, it's very interesting. So first of all, I start kicking off some work and then I go to bed. And then my designer who's in Indonesia wakes up and starts messing with Dev in. It starts prompting Dev in with red lines on annotations, which is something that Steve Louise, one of our speakers from yesterday, does with TL Draw. And I never taught him to do this. And there's no instruction manual. It was just mostly like, how would you communicate with another human being? And so I work mostly for non-technical team. And I think that's very important that they need to be comfortable with agents. And I think they're finally at the point that they are. We start working on things that we would never normally have worked on. Nobody has supported this, so I assume none of you had discovered it. But there's the Easter egg on the website. Why? Because I put it there. Why? Because it was fun. Because I could. Right? So if you're on an ultra-wide, you scan your mouse over the highlights, you see an Easter egg. I saw a tweet that was viral about a design aesthetic that I liked. I threw it into Dev in. Out pops. And then 127 replies later. Literally, I popped in and I was like, let's just see what the Clanker will do for me. I don't want to waste my designer's time. I just want to see what Clanker does for me. Designer jumps in. And it actually starts working on this thing, which I thought was throw away and fun. And the most interesting thing is so small. I can't even read it. I'm so sorry for this. So basically, the reason he starts working on it even though it's a throw away project is because it's fun. And I think that's something that was a big aha moment for me. I am getting more work out of my employees because they enjoy doing it. Because the feedback cycle for them, from like waiting blocking on me or a contractor developer that we have is gone. They literally have the idea that they go do it. And they're doing more things. They're doing animations. They're doing polish. Things that we've just, I'm getting work that I've never gotten out of my employees before. I think that's something that's appreciated. That's something you should appreciate too. If you haven't noticed, I'm no longer talking about agents for coding or like how many lines of code I'm producing. I'm getting more productivity out of my humans. And I think this is something that is a major theme for this year that I'm really trying to investigate, which is agents for everything else. Then obviously, okay, I had that success with Figma to website. I have this success with tweet to website. What else? You start to think about other use cases. This whole conference is a giant data management problem. I have to sync with 130 speakers and a couple dozen sponsors and all the attendees that come in with all various needs. And really, it's just a CMS. We've met with the sanity. I'm not the biggest sanity in the world because I want to keep some sanity to myself. But basically, I can throw in spreadsheets and Devon can manage that for me. And once I really, I think the unlock happened when I threw away the CMS and just committed that to code, but used that code as my source of truth. And Devon, whatever coding agent you use, start to manage it. And so this entire schedule is managed by Devon. What does that mean? It means that whenever someone comes in with a speaker change, for example, Marder wanted the speakers from today, sent in an email, I just say, Devon, handle it for me. No other further communication is needed. I can just forward the email. I can paste the screenshot whatever. And that kind of volume that's us as a small team of nine people manage a thousand person conference. We're going to manage 6000 people in San Francisco this summer. And I'm pretty sure we can stay the same size. It is incredible the amount of productivity that you can get once you're sufficiently onboarded and you have the workflows ironed out. We have agents for ETL. We do a external vendor system that has data that we don't have in the central source of truth. So I need to get the API key to sync over data and make sure there's a single source of truth. These are very boring routine tasks. Well, there's another fun story that I can tell you is agents for buying. So I saw this viral tweet about how somebody put a claw in Wall Street next to the Wall Street Bowl. And I was like, oh, that's funny. We should put a claw in front of our conference. And that's exactly. So I asked Devon to research, where can I get a lobster in London? Devon comes back with phone numbers and email addresses and websites. And I just clicked through and think about it and ask it to do some more research. And I'll pause this guy. That's literally the lof said that you had was bought from Devon. And I think this kind of personal automation for everything else, it just matters that you have an agent that has web access, that has some smart enough model. I mean, this is effectively a claw, right? Like an open claw, nano claw, whatever the anchor you call it, it doesn't really matter. It matters that you're using agents for things that you would otherwise have spent knowledge work on. I might have had an executive assistant. I might have had a junior employee do these things for me, but now I can do it serverless on demand with a coding agent. I'm not here to only show Devon. I just advised for the company now, but I started exploring town because I think what's happening here is coding agents breaking containment. There's all these other more fit for purpose, knowledge management tools. Like the wikis that Andre Kaprafi is talking about, that open claw is not adopting as well. You're going to see an explosion on this this year. This is probably the top trend of maybe three to five trends of 2026 that I wanted to alert you to. So here is me managing the world's fair in 2023 in this summer. Here are all the tracks I'm planning. Here's my Apple notes. On the left is my Apple notes of all the people. It's intentionally small. I threw it into town and outpops a nice, reformatted notion doc with research on all the speakers that I intend to solicit and think about curating. Then obviously, once you get enough psychosis, you are thinking about replacing entire pieces of SAS. Here is me arguing with my employees about kicking out a SAS to and building it ourselves because we can. So I clearly have the most psychosis. The thing one of the annoying things is if you are in a position of power of management to deal with employees who are not as much in psychosis and try to bring them along the journey and not talk down or ignore their concerns because they are very valid concerns because they are exactly the people that will have to deal with your bullshit when you get it wrong. We do get it wrong. One top one method I am approaching the AI replacing SAS concept, which I think should be relevant for a lot of you is let's identify the top three concerns and let's systematically reduce them and that's the process that we're going through right now. So I just wanted to give you a little bit of that taste of like here's how AI is changing our business as managing the conference. It's come really a long way. It's a consistent theme I'm seeing even among our speakers. This is multi-opening keynote talking about how the 60% of the user base of Risselle now is bots, it's agents, it's not humans. So actually your dashboards don't matter, your APIs matter, your CLIs matter, your MCPs matter. Here is the MCP apps guys Edo and Lead who spoke today on speaking on ETN about how basically your custom UI is going away, you should ship UI to somebody else's app. I think this pattern of how your primary user is changing is really shifting towards what people are calling agent experience. That's something that I'm really inspired by and focused on because it is helping me. I no longer care about the Figma dashboard, I throw it into Claude core work and I hope that it works for me. So that's my message, it just for everything else that's coming, wake up, use it, bring it home to work if people are insufficiently bot on, prescribe them one of these. Thank you.
TL;DR
- AI agents are rapidly transforming workflows beyond just coding, enabling small teams to manage large-scale operations and achieve significant productivity gains.
- The speaker highlights that AI's declining cost curve makes its widespread adoption crucial, fostering "tiny teams" with high revenue per employee.
- Crucially, agents free up human creativity and productivity by automating tedious "yak shaving" tasks and even managing complex data, shifting focus from code output to human output.
Takeaways
- The cost of AI continues to drop dramatically, roughly 100 times every 12-18 months, making its increased usage highly advantageous.
- "Tiny teams," defined as having more millions in revenue than the number of employees, can be built and scaled by leveraging AI agents for various tasks.
- AI agents are particularly effective at eliminating "yak shaving" – the tedious prerequisite tasks like installing dependencies or fixing environment issues.
- They can rapidly convert design mockups (e.g., Figma) into pixel-perfect, functioning websites, speeding up design-to-development cycles.
- Non-technical team members can effectively use agents by communicating in natural language, leading to increased productivity and even creative "fun" projects.
- AI agents can manage complex data problems, such as conference schedules, speaker changes, and sponsor coordination, acting as a flexible CMS replacement with code as the source of truth.
- Personal automation for knowledge work, like researching vendors or making purchases, can be offloaded to agents with web access.
- The rise of agents is shifting the focus from traditional UI dashboards to APIs and CLIs, as the primary user of services increasingly becomes other bots or agents.
- To adopt AI for replacing existing SAS tools, identify and systematically reduce the top three concerns associated with the change, while addressing employee concerns.
Vocabulary
Tiny Teams — Teams that generate more millions in revenue than their total number of employees, emphasizing high efficiency.
Yak Shaving — The process of undertaking a series of seemingly trivial tasks that are prerequisites to a larger, more important task.
Coding Agents — AI tools capable of generating, executing, and debugging code, often to achieve a specified goal.
Pixel-perfect — A design or implementation that precisely matches the visual specifications, down to every pixel.
Red lines — Annotations or markings on a design showing precise measurements, spacing, and other details for developers.
ETL — Stands for Extract, Transform, Load; a process in data warehousing that involves extracting data from sources, transforming it into a usable format, and loading it into a destination.
CMS — Stands for Content Management System; a software application used to create and manage digital content.
Open Claw — A reference to agent architectures (like Open Assistant or similar frameworks) that aim to provide open and extensible AI agents.
Agent Experience (AX) — A concept referring to how AI agents interact with systems, data, and human users, analogous to User Experience (UX).
SAS — Stands for Software as a Service; a software licensing and delivery model in which software is licensed on a subscription basis and is centrally hosted.
Transcript
The future of work has many paths. Our next presenter will discuss the path that he walked with Devon as he organized this very conference. Please join me in welcoming to the stage the co-founder of AI Engineer Conferences Swix. Hi everyone. I am not the chief air officer of the UK. Unfortunately he had to leave for personal reason. Thanks for staying so long. I hope everyone has a good time. Thank you. It's so enduring and heartwarming to hear from you guys. I'll take you a little bit into how we build AI Engineer with AI. It's probably the biggest revelation that I've had. So we've had a lot of really warm reception from you guys and I think it's really great. I think this is something that we really try to engineer and hopefully this is our first event in London. Hopefully you have us back next year. One thing I wanted for those who are newer to us, I do one of these key nodes every single AI. The first very first one three years ago, I talked about the productivity gain that you get from the increased usage of AI. The second one, we talked about how you should just use more AI because the cost curve of AI is going down roughly a hundred times per every 12 to 18 months and I think it's still continuing to trend that way. The third year we started to talk about tiny teams which is basically this definition that I had that teams with more millions in revenue than number of employees. I even curated an entire track at the world's fair about this where we sort of summarized it as the tiny teams' favorite if you're interested in building that. The reason I liked this emphasis is because I think people are maybe too egotistical about looking at, look for the one person billionaire or unicorn founder. Every company can have a tiny team whether you're smaller large. I think when I look at how we run AI Engineer me being the leadership of Ben Lee and myself, we are also a tiny team. This is us. It's just a nine full-time people and we are running a business that is more than $9 million so we are a tiny team. I wanted to show you the most significant changes in our workflow since we started this two years ago. By the way, this is our take in the AJIP Hill moment. Did you guys get the AJIP Hills? Yes. Very proud of this is my brainchild. If one of your co-workers is not sufficiently AJIP Hill, you should prescribe one of this. You are all AJI doctors now. Okay. Our stack was very stable and completely non-AI which is very ironic for an AI conference. We do Figma, React, SuperBase, Tido, Google Sheeds, Session and so on. I had this funny weird moment where I joined cognition and started using coding agents seriously at work mostly because they were free. I started adding it to the company's lack and then I started doing things that didn't show people, hey, here's how you use it to do coding on the company website. Oh, well, I'm good. Something strange starts happening. I started introducing this is a workflow of our contract designer in awful time showing me a Figma page and asking me to go through it and expecting that we would take a week, two weeks, four weeks to turn it into reality. I just added Dev into it and ultimately, before I had to add Dev into it, I had to cook up Dev into Figma and I'm not going to do that shit. So, core is doing it for me. You should use core for doing this. And which by the way, leads me to my first lesson, which is anytime there's random yaks shaving, I think one under-appreciated benefit of agents is that they save you the yaks' shapes, all the dependency tree crawling of like, oh, no, I have to do that first. No, I have to do that first. Particularly when it comes to installing dependencies of fixing pipe on dependencies, fantastic for that. And I think model productivity that doesn't sufficiently appreciate parallelism and not just autonomy, I think, and as a death of the yaks' shaving is not fully capturing the benefit of agents. So anyway, back to the agent story, hooked up Dev into Figma and in very short order, we have a perfectly functioning website that is pixel-perfect to the Figma. And to me, that was the surprise I've never done it before. You always mistrust marketing until you see it for yourself. And more importantly, our design knows very happy about it. And that's basically the website that you see left today when you go to AI.engineer. The other interesting thing that happened was that we started using it more, after one initial success, you started using it more. Something that you can't see because it's very small text, but I'm going to highlight for you is that that is 207 replies, just exploding in usage, like, what the hell? And when you dig into it, it's very interesting. So first of all, I start kicking off some work and then I go to bed. And then my designer who's in Indonesia wakes up and starts messing with Dev in. It starts prompting Dev in with red lines on annotations, which is something that Steve Louise, one of our speakers from yesterday, does with TL Draw. And I never taught him to do this. And there's no instruction manual. It was just mostly like, how would you communicate with another human being? And so I work mostly for non-technical team. And I think that's very important that they need to be comfortable with agents. And I think they're finally at the point that they are. We start working on things that we would never normally have worked on. Nobody has supported this, so I assume none of you had discovered it. But there's the Easter egg on the website. Why? Because I put it there. Why? Because it was fun. Because I could. Right? So if you're on an ultra-wide, you scan your mouse over the highlights, you see an Easter egg. I saw a tweet that was viral about a design aesthetic that I liked. I threw it into Dev in. Out pops. And then 127 replies later. Literally, I popped in and I was like, let's just see what the Clanker will do for me. I don't want to waste my designer's time. I just want to see what Clanker does for me. Designer jumps in. And it actually starts working on this thing, which I thought was throw away and fun. And the most interesting thing is so small. I can't even read it. I'm so sorry for this. So basically, the reason he starts working on it even though it's a throw away project is because it's fun. And I think that's something that was a big aha moment for me. I am getting more work out of my employees because they enjoy doing it. Because the feedback cycle for them, from like waiting blocking on me or a contractor developer that we have is gone. They literally have the idea that they go do it. And they're doing more things. They're doing animations. They're doing polish. Things that we've just, I'm getting work that I've never gotten out of my employees before. I think that's something that's appreciated. That's something you should appreciate too. If you haven't noticed, I'm no longer talking about agents for coding or like how many lines of code I'm producing. I'm getting more productivity out of my humans. And I think this is something that is a major theme for this year that I'm really trying to investigate, which is agents for everything else. Then obviously, okay, I had that success with Figma to website. I have this success with tweet to website. What else? You start to think about other use cases. This whole conference is a giant data management problem. I have to sync with 130 speakers and a couple dozen sponsors and all the attendees that come in with all various needs. And really, it's just a CMS. We've met with the sanity. I'm not the biggest sanity in the world because I want to keep some sanity to myself. But basically, I can throw in spreadsheets and Devon can manage that for me. And once I really, I think the unlock happened when I threw away the CMS and just committed that to code, but used that code as my source of truth. And Devon, whatever coding agent you use, start to manage it. And so this entire schedule is managed by Devon. What does that mean? It means that whenever someone comes in with a speaker change, for example, Marder wanted the speakers from today, sent in an email, I just say, Devon, handle it for me. No other further communication is needed. I can just forward the email. I can paste the screenshot whatever. And that kind of volume that's us as a small team of nine people manage a thousand person conference. We're going to manage 6000 people in San Francisco this summer. And I'm pretty sure we can stay the same size. It is incredible the amount of productivity that you can get once you're sufficiently onboarded and you have the workflows ironed out. We have agents for ETL. We do a external vendor system that has data that we don't have in the central source of truth. So I need to get the API key to sync over data and make sure there's a single source of truth. These are very boring routine tasks. Well, there's another fun story that I can tell you is agents for buying. So I saw this viral tweet about how somebody put a claw in Wall Street next to the Wall Street Bowl. And I was like, oh, that's funny. We should put a claw in front of our conference. And that's exactly. So I asked Devon to research, where can I get a lobster in London? Devon comes back with phone numbers and email addresses and websites. And I just clicked through and think about it and ask it to do some more research. And I'll pause this guy. That's literally the lof said that you had was bought from Devon. And I think this kind of personal automation for everything else, it just matters that you have an agent that has web access, that has some smart enough model. I mean, this is effectively a claw, right? Like an open claw, nano claw, whatever the anchor you call it, it doesn't really matter. It matters that you're using agents for things that you would otherwise have spent knowledge work on. I might have had an executive assistant. I might have had a junior employee do these things for me, but now I can do it serverless on demand with a coding agent. I'm not here to only show Devon. I just advised for the company now, but I started exploring town because I think what's happening here is coding agents breaking containment. There's all these other more fit for purpose, knowledge management tools. Like the wikis that Andre Kaprafi is talking about, that open claw is not adopting as well. You're going to see an explosion on this this year. This is probably the top trend of maybe three to five trends of 2026 that I wanted to alert you to. So here is me managing the world's fair in 2023 in this summer. Here are all the tracks I'm planning. Here's my Apple notes. On the left is my Apple notes of all the people. It's intentionally small. I threw it into town and outpops a nice, reformatted notion doc with research on all the speakers that I intend to solicit and think about curating. Then obviously, once you get enough psychosis, you are thinking about replacing entire pieces of SAS. Here is me arguing with my employees about kicking out a SAS to and building it ourselves because we can. So I clearly have the most psychosis. The thing one of the annoying things is if you are in a position of power of management to deal with employees who are not as much in psychosis and try to bring them along the journey and not talk down or ignore their concerns because they are very valid concerns because they are exactly the people that will have to deal with your bullshit when you get it wrong. We do get it wrong. One top one method I am approaching the AI replacing SAS concept, which I think should be relevant for a lot of you is let's identify the top three concerns and let's systematically reduce them and that's the process that we're going through right now. So I just wanted to give you a little bit of that taste of like here's how AI is changing our business as managing the conference. It's come really a long way. It's a consistent theme I'm seeing even among our speakers. This is multi-opening keynote talking about how the 60% of the user base of Risselle now is bots, it's agents, it's not humans. So actually your dashboards don't matter, your APIs matter, your CLIs matter, your MCPs matter. Here is the MCP apps guys Edo and Lead who spoke today on speaking on ETN about how basically your custom UI is going away, you should ship UI to somebody else's app. I think this pattern of how your primary user is changing is really shifting towards what people are calling agent experience. That's something that I'm really inspired by and focused on because it is helping me. I no longer care about the Figma dashboard, I throw it into Claude core work and I hope that it works for me. So that's my message, it just for everything else that's coming, wake up, use it, bring it home to work if people are insufficiently bot on, prescribe them one of these. Thank you.