Idea

AI has made observability more strategic. The biggest shift is in what engineering now needs from it.

AI has made observability more strategic. The biggest shift is in what engineering now needs from it.

October 26, 2026Systems thinkingOpen the original LinkedIn post โ†—

๐€๐ˆ ๐ก๐š๐ฌ ๐ฆ๐š๐๐ž ๐จ๐›๐ฌ๐ž๐ซ๐ฏ๐š๐›๐ข๐ฅ๐ข๐ญ๐ฒ ๐ฆ๐จ๐ซ๐ž ๐ฌ๐ญ๐ซ๐š๐ญ๐ž๐ ๐ข๐œ. The biggest shift is in what engineering now needs from it.

That was my main reflection after two days at the Enterprise Engineering Series: Platform Engineering, Developer Experience, Observability, SRE, where I moderated three keynote panels on executive strategy, reliability under constraints, and AI in engineering work.

For years, we worked to make observability part of the whole engineering cycle. AI has brought fresh urgency to that conversation: the platforms supporting AI, observability of AI itself, and AI applied to observability. ๐‘จ๐’„๐’“๐’๐’”๐’” ๐’‚๐’๐’ ๐’•๐’‰๐’“๐’†๐’†, ๐’•๐’†๐’๐’†๐’Ž๐’†๐’•๐’“๐’š ๐’Š๐’” ๐’ƒ๐’†๐’„๐’๐’Ž๐’Š๐’๐’ˆ ๐’„๐’†๐’๐’•๐’“๐’‚๐’ ๐’•๐’ ๐’‰๐’๐’˜ ๐’˜๐’† ๐’ƒ๐’–๐’Š๐’๐’…, ๐’๐’‘๐’†๐’“๐’‚๐’•๐’† ๐’‚๐’๐’… ๐’Ž๐’‚๐’Œ๐’† ๐’…๐’†๐’„๐’Š๐’”๐’Š๐’๐’๐’”.

๐๐ฎ๐ญ ๐ฏ๐ข๐ฌ๐ข๐›๐ข๐ฅ๐ข๐ญ๐ฒ ๐š๐ฅ๐จ๐ง๐ž ๐๐จ๐ž๐ฌ ๐ง๐จ๐ญ ๐ ๐ข๐ฏ๐ž ๐ฎ๐ฌ ๐œ๐จ๐ง๐Ÿ๐ข๐๐ž๐ง๐œ๐ž ๐ญ๐จ ๐š๐œ๐ญ.

Federico Vrljicak captured the challenge well: ๐ก๐จ๐ฐ ๐ช๐ฎ๐ข๐œ๐ค๐ฅ๐ฒ ๐œ๐š๐ง ๐ฐ๐ž ๐ญ๐ซ๐ฎ๐ฌ๐ญ ๐€๐ˆ? His presentation brought that question down to practical engineering choices that require deliberate work.

Ugur Basak offered another useful perspective: ๐€๐ˆ ๐ข๐ฌ ๐š ๐ฐ๐จ๐ซ๐ค๐ฅ๐จ๐š๐, ๐š๐ง๐ ๐ฐ๐ž ๐ง๐ž๐ž๐ ๐ญ๐จ ๐ซ๐ฎ๐ง ๐ข๐ญ ๐ฐ๐ข๐ญ๐ก ๐ญ๐ก๐ž ๐ฌ๐š๐ฆ๐ž ๐๐ข๐ฌ๐œ๐ข๐ฉ๐ฅ๐ข๐ง๐ž ๐ฐ๐ž ๐š๐ฉ๐ฉ๐ฅ๐ฒ ๐ž๐ฅ๐ฌ๐ž๐ฐ๐ก๐ž๐ซ๐ž. New workloads tend to look exceptional at first, but they rarely are.

One wish from Marius Zaharia stayed with me: ๐ซ๐ž๐ ๐ฎ๐ฅ๐š๐ญ๐ข๐จ๐ง ๐ญ๐ก๐š๐ญ ๐ฆ๐จ๐ฏ๐ž๐ฌ ๐Ÿ๐ซ๐จ๐ฆ ๐ญ๐ก๐ž ๐ฆ๐ž๐œ๐ก๐š๐ง๐ข๐œ๐š๐ฅ ๐ญ๐จ ๐ญ๐ก๐ž ๐ฌ๐ž๐ฆ๐š๐ง๐ญ๐ข๐œ ๐ฅ๐ž๐ฏ๐ž๐ฅ, ๐ฐ๐ข๐ญ๐ก ๐ฆ๐จ๐ซ๐ž ๐š๐ญ๐ญ๐ž๐ง๐ญ๐ข๐จ๐ง ๐ญ๐จ ๐ฐ๐ก๐š๐ญ ๐š ๐œ๐จ๐ง๐ญ๐ซ๐จ๐ฅ ๐ข๐ฌ ๐ฆ๐ž๐š๐ง๐ญ ๐ญ๐จ ๐š๐œ๐ก๐ข๐ž๐ฏ๐ž. I couldn't agree more. The intention has always mattered more than the form, but assessing whether it was being fulfilled was much harder than checking formal compliance.

My conversations with vendors also left me reflecting on where the market is going. The space is becoming more crowded while the engineering environment grows more complex. As one panelist put it, ๐’Š๐’๐’•๐’†๐’ˆ๐’“๐’‚๐’•๐’Š๐’๐’ ๐’Š๐’” ๐’•๐’‰๐’† ๐’๐’†๐’˜ ๐’”๐’•๐’“๐’‚๐’•๐’†๐’ˆ๐’š. Technology providers need to commit to helping clients connect capabilities and achieve outcomes. That commitment demands more than a strong product.

I left thinking about the distance between seeing a problem and being able to act on it. Observability helps us close that distance when the data connects to a decision, someone owns it, and the organization gives them the means to act.

๐‘‡โ„Ž๐‘Ž๐‘›๐‘˜ ๐‘ฆ๐‘œ๐‘ข ๐‘ก๐‘œ ๐‘กโ„Ž๐‘’ ๐‘๐‘Ž๐‘›๐‘’๐‘™๐‘™๐‘–๐‘ ๐‘ก๐‘  ๐‘Ž๐‘›๐‘‘ ๐‘ ๐‘๐‘’๐‘Ž๐‘˜๐‘’๐‘Ÿ๐‘  ๐‘“๐‘œ๐‘Ÿ ๐‘กโ„Ž๐‘’ ๐‘๐‘Ž๐‘›๐‘‘๐‘–๐‘‘ ๐‘๐‘œ๐‘›๐‘ฃ๐‘’๐‘Ÿ๐‘ ๐‘Ž๐‘ก๐‘–๐‘œ๐‘›๐‘ , ๐‘Ž๐‘›๐‘‘ ๐‘ก๐‘œ ๐ผ๐‘„๐‘ƒ๐ถ ๐‘“๐‘œ๐‘Ÿ โ„Ž๐‘Ž๐‘ฃ๐‘–๐‘›๐‘” ๐‘š๐‘’.

Which decision should your telemetry be helping you make today?

#ObsSRESummit #Observability #SRE #AIObservability #PlatformEngineering

View image

All ideas