Multi-agent Workflow Using Tavily

Question

Hi. I have implemented this Multi-Agent Workflow using Tavily https://docs.llamaindex.ai/en/stable/examples/agent/agent_workflow_multi/

But it is very slow compared to Tavily's own Research Assistant on their website.

This multi-agent workflow takes around a minute to provide a report on the same topic whereas Tavily's Research Assistant does it within seconds.

I am using Azure OpenAI's GPT4o as the LLM in all agents.

Are there any bottlenecks that are present or any strategies I can use to speed up this workflow?

Logan M · Answer

this example is admittedly a little contrived, but was trying to make something simple that showed the concepts

Logan M · Answer

You can accomplish the same with a single agent

Logan M · Answer

Needing to make tool calls to save the notes, report, review (and possibly passing back to the writer after the review) will all slow things down compared to a single agent making tavily calls and writing an output to return directly to the user

Logan M · Answer

agent = AgentWorkflow.from_tools_or_functions( [search_web], llm=llm, system_prompt="You are an expert report writer that will use a web search tool to help gather information to write the report requested by the user."
)

Logan M · Answer

That will probably get the job done

Find answers from the community

Multi-agent Workflow Using Tavily