Talking Proactive Optimization, ChatGPT Issues & More

Powerful things happen when ITOps teams move beyond a break-fix approach and lean into proactive optimization. Instead of just responding to issues as they occur, when teams have independent visibility into their end-to-end service delivery chain, they can proactively identify possible areas for optimization and improvement. For example, streamlining one small part of a complex process could shave seconds off the total transaction time; do this for every part of the process, and the efficiency savings can quickly add up. In recent weeks, it appeared OpenAI’s ChatGPT may have been undergoing this type of optimization, as we observed material improvements in page load times and some evidence suggesting possible configuration changes and re-architecture in the pursuit of performance improvements. Read on to learn more about what we saw at ChatGPT, as well as insights from incidents at companies including Grammarly, Bluesky, and Netflix, or use the chapters below to jump to the sections that most interest you. CHAPTERS 00:00 Intro 00:48 ChatGPT Disruptions 08:17 Grammarly Issue 10:59 Netflix Issues 12:09 Bluesky Disruptions 13:26 Salesforce Outage 14:50 Verizon Fios Disruption 16:14 Outage Trends: By the Numbers 18:36 Get in Touch ——— For additional insights on understanding and proactively optimizing digital experiences, read this eBook: https://www.thousandeyes.com/resources/guide-to-next-generation-assurance-ebook?utm_source=transistor&utm_medium=referral&utm_campaign=fy25q2_internetreport_q2fy25ep2_podcast ——— Want to get in touch? If you have questions, feedback, or guests you would like to see featured on the show, send us a note at [email protected]. Or follow us on X: @thousandeyes