Interesting approach. Do you have some benchmarked scenario’s to back up the claims of scalability vs the regular http lambda model without the worker pool? It would be interesting to see a comparison of these approaches on certain request intervals / concurrency levels.