AI Agent Inside Basecamp Automated 95% of Routine

Over the last six months, I've automated about 95% of my routine tasks by talking to AI agents 🤖 inside Basecamp. In this video, I show exactly how it works under the hood, so you can build the same setup yourself or just copy mine. Here's what's covered: ⚡ Give an agent a goal from a regular Basecamp message, with no terminal and no new tools 👥 Collaborate with teammates in the same thread and let the agent pick up every update ☁️ Run your clankers in a secure cloud machine for every separate topic 🔎 Watch the agent research, generate images, and transcribe voice notes 🧩 Interact inside durable artifacts for brainstorming and effective multiplayer decision-making 💸 Staying provider-agnostic to keep model costs down Chapters: 0:00 Intro 0:24 Two problems to solve 2:04 Giving the agent a goal 3:00 Multiplayer threads 3:23 Under the hood 4:07 Watching the agent work 6:18 Results + room preview 6:58 Voice notes 7:59 Multi-model, lower costs 9:19 Deciding with artifacts 11:08 What's next Happy to learn more about your domains: ask me anything in the comments 👇