# Brainlabs' Get Stuff Done System with Notion ## Метаданные - **Канал:** Notion - **YouTube:** https://www.youtube.com/watch?v=MAGnwc6wjHo - **Дата:** 13.05.2026 - **Длительность:** 5:23 - **Просмотры:** 954 ## Описание Watch how Brainlabs built the foundation first, then made work move. Get the full story: https://notion.com/customers/brainlabs Brainlabs built a “Get Stuff Done” system in Notion where meeting notes turn into tasks, and custom agents take it from there. Notion AI captures the context, Workers run the reliable execution steps, and Cortex feeds insights back into the workspace so teams can move faster with less busywork. Learn more about Notion's Developer Platform: https://notion.dev Watch more: Keynote (Ivan Zhao) → https://youtube.com/live/rpE2rzKO6L0 How to Sync Any Data Source → https://youtu.be/VhE16Z1mgCs How to Build Agent Tools → https://youtu.be/66dY0WCDPTU How to Orchestrate Any Agent → https://youtu.be/zkKPchV2bEk Follow along with Notion: X: https://x.com/NotionHQ LinkedIn: https://linkedin.com/company/notionhq/ Instagram: https://instagram.com/notionhq/ TikTok: https://tiktok.com/@notionhq Threads: https://www.threads.net/@notionhq ## Содержание ### [0:00](https://www.youtube.com/watch?v=MAGnwc6wjHo) Segment 1 (00:00 - 05:00) getting into Notion, discovering agents in that space, and just enjoying custom building agents has just yeah opened my eyes to what a real AI agent is. It's not just an LLM with a response time. It's a it's one independent worker. All I had to essentially do was just describe what I needed it to do, and it would build it in the format that it needed to do for a Notion worker. So, I'm a principal engineer here at Brainlabs. I've been here for now about 5 years. We are a market digital marketing agency um operating globally. So, I would say it's definitely been kind of a year or so exploring what an agent is, but it's only I would say when we've actually adopted Notion into um our kind of day-to-day lives that we've really explored a real AI agent. In a way, we should be able to do most of our business through Notion because of our agents that are enabled on there. Cortex is our general microservice platform um that we've built internally for um our client teams to be able to use. So, there's about I think there's about 16-ish tools on there, and we basically put an agent for each of the different apps. So, I had all the knowledge. The level-up that Notion provides it is that once that data is created in Cortex, it can be sent back into Notion. So, whether that is things like slides, outputs, or data analysis that's done, cuz then that data is now available to be used for the Notion agent. So, it creates context to them, so it can you can say things like, "Hey, how's our you know, campaign X, Y, and Z doing this week? " "Oh, well, based on the data that's run from the previous audit, we're seeing a 50% uplift in click-through rate, that kind of thing. " Um whereas before, you know, we have these multiple different microservices in Cortex, but how do we actually relate them to each other? That's up to a person to understand and be able to relate to that information themselves. Whereas an agent can be set up to do that in the space of an hour. The more fun one from myself was the ability to generate images within Notion. So, we've got access to Nano Banana. We've got an API key set up within a worker that knows how to generate an image. So, for this one, I'll show you the agent basically working on this when a task is created. So, that's the task being created. The image generator has recognized that it's got a um image tag on it. So, it knows, "Okay, I'm going to be the one that works on this. " The image agent will go away and work on this. Um it will first analyze like what's being asked. And then when it starts working on the generation part of the images or whether it needs to It might need to ask questions at some point, but it'll then move it into in progress. So, it just decided it was too fast and jumped all the way to review agent work. Uh it's giving you a nice little like agent progress. So, it said it's generated the initial one, and I can go in here to review this document. And it's come up with Brain Labs X Notion. We're using our yellows and blues. Uh it's let us know that the prompt that it's used. It's also made in like noted any assumptions based on our brand guidelines. Uh and then the kind of different style uh variants and things as well. It's recognized that there's been a task created, which was obviously this one. Um it's jumped on that and basically gone through. It knows the the tools that it's got access to, which down here is the image generation. So, it's got generate image and refine image. So, obviously in this case, it's just generating an image based on um prompt and also the context it's given. And then yet it knows as well that like once it's finished set it to review agent work and that allows me to go on and be like, "Yeah, this is great. Actually, I want you to do this instead. " Or um I can then if I'm happy enough, I can just essentially be like, "Okay, do you know what? That's great. Part of my day is done. " But the biggest push for us when it comes to what we want to do with agents going forward is we want to continue to enable agents to do essentially everything. Um and not in like a we're replacing ourselves type way, but in a humans have so many parts of their day that have without even thinking about have just become parts of our day. But if we actually like boiled it down, there's probably only 30% of your day that you realize is — is part of the reason that you're in that job in the first place. And the other 70% just happens to be side products of that. And the more that we can enable agents to do those kind of won't say meaning- meaningless tasks ### [5:00](https://www.youtube.com/watch?v=MAGnwc6wjHo&t=300s) Segment 2 (05:00 - 05:00) but like tasks that just aren't using 100% of your cortex, not in the brain labs way, but in the real brain way. Um then you know, we're allowing us as professionals to be able to do more of what we love, but also more of what we're good at. --- *Источник: https://ekstraktznaniy.ru/video/50803*