Self-hosting Models and Managing Token Spend

About this event

**Register at the official event page to attend: https://san-diego.aitinkerers.org/p/self-hosting-models-and-managing-token-spend** (registration is screened based on proof of work; RSVPing here on Meetup does not reserve a spot) AI Tinkerers San Diego is back for another evening of code-first demos, messy experiments, and high-signal technical exchange. This month we're focused on self-hosting models and cutting token costs: taking AI systems from interesting demos to reliable production without letting spend get out of hand. This is a builder-to-builder forum for people shipping AI. Come prepared to show the system, explain the architecture, and talk candidly about what worked and what didn't. Leave the slides and sales pitch behind. **What to Expect** * Community Demos: short, code-first presentations. Live systems preferred. * Technical Q&A: deep discussion after each demo * Builder Networking: trade patterns, debugging tricks, and war stories * Open Demo Zone: bring a laptop and show what you're building **Agenda** * 6:00 PM Doors open * 6:35 PM Welcome and housekeeping * 6:45 PM Demo 1: Self-hosting inference * 7:00 PM Demo 2: Cutting token usage * 7:15 PM Break and food refill * 7:20 PM Demo 3: Applied AI outside traditional tech * 7:35 PM Lightning round * 7:50 PM Wrap-up and next meetup * 8:00-9:00 PM Networking and open demo zone **REMEMBER TO BRING YOUR ID.** No ID, no building access. Room capacity is 50. Hosted at the AWS San Diego office (address shared with confirmed registrants). AI Tinkerers San Diego is 100% volunteer-run by Max, Diana, and Michael. Huge thanks to AWS for supporting this event.