Hey everyone — new here, just joined yesterday. Figured I'd share what I'm working on instead of just lurking. I run Inspire Applied, an advanced prototyping shop in Massachusetts. Lasers, 3D printing, CNC, the works. Last year I started building an AI agent that handles the front end of the business — quoting, qualifying projects, routing work, logging, follow-ups, even content. The agent lives at https://makerasaservice.defyimpossible.tech/ — if you visit, you're talking to it directly. It quotes your project, qualifies the work, and when you're ready, the physical shop executes. No chatbot handoff to a human. The AI is the interface between the customer and the machines. What it does right now: - Takes project descriptions and generates quotes in minutes - Connects to Gmail, Calendar, Drive, GitHub, Notion, YouTube, LinkedIn - Logs everything, remembers context across sessions - Routes work to the right machine/process - Handles the full customer conversation What I'm working on: - Tighter integration with shop floor systems (machine status, scheduling) - Automated content generation from completed jobs - Deeper qualification logic (material selection, tolerances, cost optimization) Would love feedback from people who've built or are building persistent agents — especially ones that connect to physical systems. What's working? What's breaking? What would you do differently?