Free Lesson
Avoiding expensive AI surprises with LiteLLM
60 min
Feb 5, 2026 12:30 PM
Virtual (Zoom)
In this video
What you'll learn
Put real cost and usage boundaries around LLM experiments
use local proxies like LiteLLM instead of relying on opaque hyperscaler abstractions to manage LLM use
Trade-offs between different ways of accessing Gemini
CLI tools vs APIs vs wrappers - when each approach actually makes sense
What works and what doesn’t running LLM tooling locally
Including lessons from Nix, Apple Containers, and lightweight setups without Docker
Why this topic matters
LLMs have lowered the barrier to experimentation, but raised the risk of silent, unbounded cost. For individual engineers and small teams, the gap between “enterprise best practices” and day-to-day reality is huge. This talk focuses on reclaiming control: making costs visible, experimentation safe, and tooling choices explicit, so you can move fast without unpleasant surprises on your credit card
You'll learn from

Ewan Nisbet
Technologist

Doug Turnbull
Search Practitioner
Coach Search Teams at
.png&w=1536&q=75)