I Built an AI Media Buyer in Slack (Full Breakdown)

Download the free Paid Media OS: https://join.tinofirm.com/paid-media-ui/ This AI media buyer works inside Slack to analyze Meta Ads data, discuss campaign strategy, review creative and copy, and carry out campaign instructions. In this demo, I show the supervised workflow I built with Atlas, a Hermes agent connected to Meta Ads. Join the AI Systems Builders waitlist: https://join.tinofirm.com/ Watch the Meta app connection setup:    • My AI Agents Run Meta Ads for Me (Free Dow...   The important part is not pretending that AI replaces judgment. I set the budget, provide or review the creatives, discuss the audience and campaign structure, and explicitly authorize campaign actions. The agent helps connect current account data, operating instructions, creative feedback, campaign setup, and reporting in one workflow. What you’ll learn: → How Slack can become the interface for a Meta Ads AI agent → What an AI-assisted campaign audit can include → How the agent handles a Slack voice note and updates its analysis → How reusable skills can encode a paid-media operating process → How to discuss budget, ad sets, audiences, and campaign structure with the agent → Why sensitive-attribute language needs careful Meta policy review → How human-created creatives can be reviewed before campaign setup → What the prompted campaign-creation and stop sequence looks like → How Hermes, Slack, a Meta app, API access, and MCP fit together → Where human supervision and account-specific permissions still matter Important: this video demonstrates one supervised workflow. It does not prove autonomous performance, profitability, universal policy compliance, uninterrupted 24/7 operation, or results for every account. Use explicit account scope, review gates, paused-object creation where appropriate, and separate approval for activation or material budget changes. Chapters: 00:00 The AI media buyer in Slack 00:28 Asking Atlas to audit a Meta Ads campaign 01:30 Voice notes and current campaign data 02:32 Teaching the agent your paid-media process 03:02 Budget, ad sets, and audience strategy 03:56 Meta policy and creative guidance 05:04 Reviewing creatives and copy 05:57 Creating and stopping campaigns 06:42 How the Hermes agent works 08:30 Reporting, monitoring, and human supervision 09:01 Connecting through a Meta app, API, and MCP 09:46 Download and AI builders waitlist If you want practical videos about AI agents, company context, second brains, and operating systems, subscribe. About Gui Costa: I build practical AI systems for companies that want their context, workflows, and tools to work together. This channel covers AI agents, second brains, operational systems, and the implementation decisions behind them. #AIMediaBuyer #MetaAds #AIAgents