Merit
Sign in
DiscoverMarTechNotion
MarTech

Brainlabs' Get Stuff Done System: Custom Agents That Execute While You Focus on Deep Work

66/100validation score
70%
busywork time freed for strategic work
500units
clients managed with automated workflows
30%
time spent on strategic work (increased from busywork reduction)

Brainlabs built a system in Notion where work moves from captured context to completed execution without waiting on humans. Custom agents handle deterministic steps, connect to their proprietary Cortex platform, and complete tasks automatically—removing busywork and speeding up delivery across hundreds of clients.

By the time this meeting finishes and I go make a cup of tea, all those tasks are already done for me in the background.

Josh, Principal Engineer

I'm sending a task off to something and it's doing it in the background. But that can often just be a very delayed return on an API call.

Josh, Principal Engineer

There's probably only 30% of your day where you do the work you're meant to do in your job in the first place. The other 70% is dealing with the side products of that work.

Josh, Principal Engineer

If we decide to up and shift over to a different LLM, all of our sources of information still connect to the same place: Notion.

Josh, Principal Engineer

Brainlabs case study video

Brainlabs' Cortex AI platform

Brainlabs' Cortex AI platform

Brainlabs' Agent Library

Brainlabs' Agent Library

Brainlab's Image Generator Agent

Brainlab's Image Generator Agent

1

About the Vendor

Notion is an AI-powered workspace platform that combines docs, projects, knowledge bases, and AI agents. The platform enables teams to build custom agents that automate busywork, connect to external tools, and execute tasks in the background. Notion's developer platform allows companies to create orchestration layers where agents can access knowledge bases, execute Workers (JavaScript functions), and integrate with proprietary systems—all without managing infrastructure.

2

About the Brand

Brainlabs is an international media agency managing campaigns for over 500 clients. The company has a small, tight-knit engineering team that builds best-in-class proprietary tools in-house, including Cortex—a suite of internal tools that client teams use to optimize budgets, audit accounts, and manage campaign setup across every major ad platform. Brainlabs operates globally with teams across multiple regions.

3

Challenge

Brainlabs had powerful internal tools in Cortex, but their outputs were siloed—readable only by whoever ran the tool and invisible to everyone else. When LLMs arrived, the team layered them into Cortex quickly, but there was a gap: tools that took a task and returned an answer weren't really agents. A real agent needed to go off, work, surface questions, wait for answers, and keep going while users moved on to other tasks. The company also faced the challenge of 70% of employees' time being consumed by busywork—the repeatable, time-consuming side products of their actual work—rather than the strategic 30% they were hired to do. Client onboarding involved long checklists of setup tasks spread across various tools in Cortex, still requiring manual management. Additionally, Brainlabs wanted employees across every job function and region to develop and share AI skills without requiring technical expertise or touching a terminal.

4

Solution

Notion became Brainlabs' visibility layer and orchestration interface for their entire company and proprietary Cortex platform. The solution includes Notion Custom Agents that automatically pick up and process action items, Notion AI for meeting notes and task generation, and Workers (JavaScript functions) that execute deterministic actions reliably. Notion serves as the central knowledge base where skills, context, and task history live—accessible to agents regardless of the underlying LLM or platform. The platform connects to Brainlabs' Cortex system, allowing agents to pull insights and push results back into the workspace. Notion also provides a PR-style review system where employees can submit skills that get formatted, categorized, reviewed, and pushed to GitHub automatically—all without writing code directly.

5

Execution

Josh Reid, Principal Engineer at Brainlabs, built the system starting with client onboarding as the first proof point. When a client gets created in Notion, an agent reads the page, spins up a database of tasks, and starts completing them automatically before the account team touches anything. Josh built a PR-style review system in Notion where employees submit skills, an agent formats and categorizes them, approvers review in Notion, then everything is pushed to GitHub and back into Claude automatically. He implemented Workers—JavaScript functions that define exactly how actions should execute—and handed them to agents as tools. The agent reasons about when to use them while the function handles execution deterministically. An image generation agent demonstrates the full system: it picks up tasks, generates multiple options using a Worker that calls an image API, and drops them into Notion pages for review, already knowing Brainlabs' brand guidelines from Notion. Josh built Workers while barely writing code directly—he just tells Claude what he needs.

6

Results

By the time Josh finishes a meeting and makes tea, all tasks generated from meeting notes are already completed by agents working in the background. The system removes the 70% of time employees previously spent on busywork—the repeatable, time-consuming side products of work—allowing them to focus on the strategic 30% they were hired to do. Client onboarding tasks that were previously spread across tools and required manual management now complete automatically before account teams touch anything. The architecture ensures that if Brainlabs decides to shift to a different LLM, all sources of information still connect to the same place: Notion. Employees across every job function and region can now develop and share AI skills without technical expertise, using the PR-style review system built in Notion.

Tech Stack

CortexClaudeGitHubSlackImage API

Validation breakdown

Audited against Merit's standard
Evidence quality80/100

How well the outcomes are documented and traceable.

Authority signals46/100

Brand recognition, client relationship, and external references.

M
Merit Validation Team✓ Verified
Independent audit · Merit standard

Data source: declared · mention