Build Smarter AI Agents for Free Using LangGraph & Ollama

How to create agentic systems from scratch and monitor them using LangGraph studio?

Image by Writer using Canva

Imagine you have a chatbot, but it’s not very smart. This means that you provide input to your chatbot, and it generates a simple response using an LLM.

The good news is that you can turn your chatbot into an agentic system simply by integrating AI agents into it.

Cool, right?

Let me say that the concept of the agent is not new; it has existed in the world of AI before. It is defined as “one that acts or exerts power”.

Let’s take a closer look at the concept of “act.”

Every act involves a decision. When it comes to decision-making, there must be multiple options. To select the best option, you need access to accurate and up-to-date information.

This is exactly where the AI agent comes into play.

AI agents evaluate various options, reason through them, and choose the most suitable answer based on available data. To do this, they leverage LLMs and, when necessary, external tools.

The agent uses a loop to carry out the task you give.