Challenges that I'm currently working on in the development of AI Agents
So here are 3 problems that I'm working to solve in the development of our agentic systems.
Any help would be truly appreciated.
1) How to prevent the agent from leaving the loop in between tool calls, saying it's "going to move on to the next step", or to ask the user for confirmation.
2) How to make the agent manage it's tokens autonomously and intelligently, by identifying and discarding messages or context that are no longer relevant.
(Important in particular for when there's this one gigantic piece of content among the messages that keeps getting sent back in each new API call, making the token count grow exponentially)
- One path that I'm exploring is to have the one task or step in a different agent, so the main agent calls the agent that executes this task, so the context stays separate.
3 - On that last note, to understand the balance between having 1 agent with a huge prompt and a lot of tasks/steps, or one "manager" agent that calls multiple small agents with small prompts and single tasks.
Replies
Hi Rodrigo,
Can I ask what framework you are using? I ask because just yesterday, my colleague shared their findings on the difference results they got using smolagents vs using Agent SDK (Open AI). Our team is evaluating what tool to use before we get into building the agent system.
@mubashirullahd Actually I'm not using any framework, or tool for that matter. It's more like I'm building one my own, but that's just due to the way I want to have the agents integrated to our product.
I really wouldn't know what to recommend you as far as tools go. If I had to blindly guess though, I'd totally go with OpenAI. Again, I have no idea what AgentSDK is or how it works, but one thing I can tell you after obsessively tinkering with their stuff for a while now, is that these guys are devoted to provide other engineers with the very same same tools that THEY use to build their products. So a good bet for the long run as well.