The node-image cutover protects the cluster while shrinking the margin that absorbs peak inference traffic.
The security deadline requires the old image to leave service, while researchers need answer requests to survive backend turnover.
Who you’d be doing this for
“I had the sources open, then the answer just froze while my client was waiting.”
Yunho Han · Independent financial researcher
A repeat consumer researcher using Search to verify fast-moving market information against cited sources.
What is at stake
Some answer requests stop halfway through whenever inference nodes drain during peak research traffic. You have to weigh removing the exposed node image against preserving capacity for live searches.
Why it isn’t already fixed
Every obvious fix costs something else. That’s the part you’d have to decide.
- security closure vs answer availability
- capacity certainty vs AWS cost
- fast cutover vs production verification
- backend readiness vs Kubernetes health
Why Perplexity
Perplexity Search depends on Kubernetes inference capacity to return source-cited answers while web research demand changes by the minute.
Written with these in mind
Not your kind of problem? 31 more at Perplexity, or browse every organization.
This is the setup. The work is inside.
Running it puts you in the room: the full situation and its constraints, stakeholders who push back in their own words, and the decisions that are yours to make. What you produce becomes a Day One Plan — work you can show someone instead of describing.