One-Day Intensive
Control AI Spend + Hosting Open Models: 1-Day Intensive
10.17.26 | 10:00 AM - 4:00 PM PDT 6 hours in one focused day
Own the infrastructure your AI runs on, locally, on your cloud, and at scale across every service you ship.
with Liz Howard
Overview
Every AI feature you ship sits on somebody's inference infrastructure. This is the day it becomes yours. We provision it end to end, the tooling, the hosted platforms, your own cloud, your own hardware, and the laptop in front of you, and then push it until it serves many services at once without falling over. You will leave with a running OpenAI-API-compatible endpoint, a client that cannot tell it apart from a commercial provider, and a costed argument for which one you should actually be paying for. Bring a machine with a dedicated GPU or an M-series Mac with 16+ GB of RAM. No GPU is fine, you will provision cloud instead, and we will cost that honestly too.
Prerequisites
- You can use a terminal or command line (running commands, editing config files, SSH) - You can read Python code, you will configure and run existing code, not write it from scratch - You know what an API is and how to call one - Helpful but not required: Docker, cloud hosting (AWS/GCP/etc.), or Linux system administration - Hardware for local hosting: a dedicated GPU (NVIDIA recommended) or an M-series Mac with 16+ GB RAM, or use cloud hosting instead - Costs: cloud hosting runs roughly $0.50–5/hr depending on GPU; self-hosted is free after hardware
What You'll Master
- Oobabooga and LangServe as OpenAI-compatible backends
- One client, two backends, swap only the base URL
- Provisioning on your own hardware, your own cloud, and HuggingFace
- postmarketOS phones as $0 inference nodes
- Cloudflare Tunnel and Access, no port forwarding, no static IP
- Routing and fallback across many services at once
- Cost per outcome, and the traffic level where the answer flips
Where This Class Leads
What you leave able to do
- run an open model locally
- configure an open model endpoint behind a provider interface
- reproduce your environment on a machine that is not yours
- classify models by fit for a workload
- justify a hosting choice by cost per outcome
- originate a replacement stack when the economics change
How you show it. A cutover record with cost per outcome before and after, the compatibility surface that kept callers unchanged, and no interruption in the traces.
Join Us
One-Day Intensive
- Your seat in the class
- Lifetime access to materials
- Sponsors 1 scholarship seat for someone else
The minimum is $350, it keeps out spam. Need to pay less? Apply for a scholarship, no one is turned away for lack of funds.
Secure payment via Stripe
One-Day Intensive
- Your seat in the class
- Lifetime access to materials
- Sponsors 1 scholarship seat for someone else
The minimum is $175, it keeps out spam. Need to pay less? Apply for a scholarship, no one is turned away for lack of funds.
Secure payment via Stripe
One-Day Intensive
- Your seat in the class
- Lifetime access to materials
- Sponsors 3.3 scholarship seats for someone else
The minimum is $750, it keeps out spam. Need to pay less? Apply for a scholarship, no one is turned away for lack of funds.
Secure payment via Stripe
No sponsored scholarship seats are open in this class yet. Not enrolling, but want to put someone through this class?
Instructor
Want to schedule this class for your team?
Contact us: liz@themultiverse.school