I have built a crew with a single agent and a single task. Uses 2 tools, search and scrapper. The LLM inference itself is fast however i see it spend a lot of time at certain instances during execution (iterations) eg. after doing a search, it takes a lot of time sometimes to move ahead. Also after errors. I am not using memory and i havent set any limits on the number of calls or iterations.
Is there a way to speed up the entire process.