This is impressive and I'd love to try it for our logs. Our company has specific requirement about resources and ranking but I couldn't find it in documentation

This is impressive and I'd love to try it for our logs. Our company has specific requirement about resources and ranking but I couldn't find it in documentation, could you please point me to the right docs? 1. How does it scale with increased search traffic? What are CPU and RAM requirements for 1'000 request/sec? What if this is just spike traffic, and average is 1 request/sec? 2. Proximity / positional search - our current workaround is to move log entries with tokens closer to each other higher in the results with time penalty quadratic function, but ideally we'd like proximity operators that would eliminate results with tokens far away.