A very complex set of conditions occurred across numerous systems that was legitimately unpredictable. A system I supported looked like it was giving back false data in a seriously bad way that may have resulted in inaccurate data / data loss that may have been undetectable for a long time ... catastrophic if true.

All the evidence was from some testing they did, but it lined up with what was seeing as "bad data" in a way that seemed rock solid.

Despite all the evidence from their testing it just made no sense to me that what they claim was occurring was in fact occurring based on my experience, and what my own engineering folks I worked with said.

So I proposed a test that was pretty laborious but also would absolutely answer a lot of questions. The banking folks being good guys were all for it as they were inundated with executive pressure as there was a real possibility that there were regulatory implications, possibly enormous.

So anyway we run the test and everyone reports what they see along the chain of networking equipment and data and so forth. And it comes back completely unexpected nearly every step of the way. It also clearly points to what was a fundamentally bad design an some misunderstandings all along the way.

The conference call ended with the bank "We're going to a breakout session with the executive team, please stand by." A few hours later they told us we could go.

Thankfully no data damage was done, it was possible someone really fucked up as far as regulatory stuff but it really depended on your POV, and the events were highly unusual / unpredictable. No idea if the bank ever reported it to the feds.

Nobody talked about it after that, just as you noted, everyone pretended nothing happened.