PastAGI Weekly

Practical AI engineering, audited

Slide 1

Weekly guides and audits for engineers building with LLMs: serving-cost math, eval methodology that catches silent failures, and hardware analysis with numbers you can reproduce. Every claim traces to a benchmark or a receipt.

Sent Daily Written in English
SubscribersLock iconLock icon
Open RateLock iconLock icon
Click RateLock iconLock icon
Added OnLock iconLock icon
Sam Racicot

Featured Issue

Close buttonClose button

KV Cache Math for Million Token Agent Runs

KV cache math explains why every decoded token reads the whole context, and what eviction and quantization can cut from million-token agent bills today.

Read Past Issues Close buttonClose button

Check out PastAGI Weekly's archive of previous issues of the newsletter.

Tools

Stats

See how this newsletters subscribers and engagement have changed over time.