Koratana: Traditional SaaS records final outputs but loses decision context
Animesh Koratana · Great Founders Aren't Writing Code. They're Building Its Immune System. | Animesh, PlayerZero · May 20, 2026 · at 11:19
Animesh Koratana, founder of PlayerZero, discusses why existing enterprise SaaS platforms fail to capture institutional reasoning.
0:00 / 0:29exact quote · 29.5s
▶ Watch the full episode on YouTube → 720p mp4 · rendered on demand · StarZero watermark
“The key insight behind context graphs was that the last decade of SaaS basically has been built around this idea of a system of record, right? I think Salesforce is a good example, right? Probably the most, you know, central system of record, probably the most popular one right out there. Salesforce is a great system of record, but it represents the output. It represents that this deal was sold for this price at this time. But a lot of the decisions and the path to getting to that deal ends up being lost.”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →
More from Animesh Koratana
Koratana: Enterprise AI agents failed to deliver on expectations in 2025
“And the last year, right, 20, 25 was supposed to be the year of agents, especially the year of agents in the enterprise. But it never really delivered.”
Koratana: SRE, QA, and support engineering functions will merge over time
“I actually, I genuinely think that as we start centralizing this context, all of these different functions that we were talking about earlier, right? Starting from SRE on like the most, you know, reactive, you know, infrastructure heavy side all the way down t…”
Koratana: Current coding agents are not optimized for remediation
“Today I don't think the coding agents are actually optimized well for that at all.”
Koratana: AI models enable autonomous agents to run entire workflows
“I think the really interesting inflection we've seen over the last three or four months is that the models have gotten good enough that agents can actually run the entire workflow, right? They're not additive to individuals. They are they're the ones that own …”
Koratana: GitHub Copilot originally began building on a GPT-2 Codex variant
“In 2019, 20, 20 time GPT-II had just been trained. And that was, I think, one of the first language models that came together and was coherent over longer horizons. And there was a variant of GPT-II that that was called Codex. And this is kind of the same name…”
Koratana: Understanding software is the bottleneck to production engineering and support
“How do you understand software and how do you model it? Because understanding often tends to be the bottleneck to being able to, You know, inflect change and to be able to do all the support and production engineering.”