Project: [general-use-gemini-cli-etc]
Account holder: Corey Slick / Slick Engineering & Consulting, Inc.
Intended use: I run a personal/business research-automation pipeline that produces weekly digests covering optics & signal-processing papers, AI tooling advancements, embedded-AI hardware, and related engineering domains. I’d like to programmatically turn each digest into a NotebookLM Audio Overview (MP3) for offline listening.
Volume: Low — roughly 5–10 podcasts per week, each generated from a single multi-source notebook (~10–50k tokens of input). Single-developer use, no end-user-facing routing.
Why the API vs. the consumer UI: I’ve prototyped against the consumer web app via Playwright but it’s UI-fragile and has no SHORT/STANDARD or focus-prompt control. The official Podcast API endpoint at discoveryengine.googleapis.com/v1/projects/.../locations/global/podcasts is the right long-term integration point.
Existing GCP usage: Active project with Discovery Engine API enabled, plus Analytics, Search Console, and Indexing APIs in production for slick-engineering
Thanks for considering — happy to provide additional context if helpful.