“Prompt bloat… there was an error in the code, it kept causing it to loop… retries reused all those tokens again… they changed it and it was off the list”

Jason Ward · Agentic Conversations (formally mlops.community) · How a Logistics Giant Keeps AI Data Locked Down

Play from 9:24
9:24

Audio streams from the publisher’s own feed. Verbatim does not host or copy it.

What was described

identified a high-prompt-bloat service, traced it to a code error causing retry loops, and fixed it to eliminate wasted tokens

Said nearby on the same subject

“If there’s a problem … we’re going to fix it right then … we have this feedback loop … we’re going to fix that week.”
NC
Nick Coleman
Lawyerist Podcast ·
How they work Share this moment
“They're using that spill error now. So it will just cause a spill, which is quite frankly, the day Dynamic Arrays showed up was one of the best updates ever.”
KS
Katelyn Stienen
FP&A Today ·
“We identified 10,000 broken links or dead web pages that needed to be fixed and it fixed them in like a day. That would have been 5,000 human hours to go through and find all of those and fix it.”
AO
Aaron Olds
The CMO Podcast ·
Automation 10,000 broken links; fixed in a day; saved 5,000 human hours Share this moment
“‘We’re using other models for static analysis because it’s a waste of tokens…’”
OS
Omar Santos
Security & GRC Decoded ·
Why it worked Share this moment

Moments like this, every Monday

The strongest new moments on your topics, in the speakers' own words. Free for three topics.

No card. No password. One-click unsubscribe.

Track a competitor?

Pro watches up to 25 names or phrases and emails you the morning after they are said.

See Pro, $29 a month

Verbatim quotes a short extract and links to the episode; the episode belongs to its publisher. Remove a show.