Create a conversational agent with LlamaIndex in 5 minutes

Tech Lead, CTO AXI Technologies

Romain DE LA SOUCHÈRE

Published on 26 mai 2025 · 10 min of reading

Table of content

Agents in LlamaIndex

The problem: LLMs cannot calculate

The solution: an agent with a custom function

And the ReAct agents?

Conclusion

Share with

LlamaIndex is an open-source Python framework designed to quickly build applications powered by LLMs (Large Language Models). Specializing in the management and querying of unstructured documents, LlamaIndex also enables the creation of intelligent agents capable of executing custom functions.

Since version 0.12+, LlamaIndex offers simple-to-set-up agent workflows that enhance the capabilities of LLMs with external tools.

Agents in LlamaIndex

An agent is a conversational system that can:

Understand a user query
Identify whether a tool or function should be called
Dynamically call these functions and interpret the result
Respond to the user with the correct information

This allows us to surpass the limitations of an LLM alone.

The problem: LLMs cannot calculate

Let's take a simple example. If we ask an LLM to perform a complex multiplication:

python

The expected answer is: 121234 <span data-md-literal="*"></span> 4561237 = 552977006458

The answer is likely to be incorrect. Why? Because LLMs are very good at predicting words, but not at performing precise calculations on large numbers.

The solution: an agent with a custom function

With LlamaIndex, we can create an agent that calls a Python multiplication function.

python

This time, the LLM understands that it needs to call the multiply function, and you get the correct answer.

And the ReAct agents?

The functional agent approach is simple and effective; it resembles the operation of ReAct (Reasoning + Acting) agents, a powerful method combining step-by-step reasoning and tool calling.

These agents are capable of devising a strategy before acting, making them ideal for complex cases, such as navigating through documents, planning, or multi-step interactions.

Compare the LlamaIndex approach with the approach from LangGraph, find our similar article on creating a ReAct agent with LangGraph: Build a ReAct Agent using LangGraph.

Conclusion

LlamaIndex allows you to create powerful agents in just a few lines of code that can enhance the performance of LLMs, especially for specific tasks such as calculations, database queries, or API calls. In less than 5 minutes, you have a reliable and extended conversational agent. Ideal for integrating AI into your professional applications.

Want to go further?

This topic is part of our Generative AI for Developers course. Browse the full programme, or get it by email.

Share with

Romain DE LA SOUCHÈRE

Tech Lead, CTO AXI Technologies

Expert Data Engineering et Cloud, Romain affiche plus de 11 ans d'expérience, dont plusieurs années comme Lead Developer sur des solutions Smart Building haute performance. Il y a conçu et mis en production des moteurs de traitement capables d'absorber des centaines de milliers de données de capteurs par minute, ainsi que des bases clusterisées gérant plus de 10 millions de données dynamiques. Certifié Microsoft Azure DevOps Engineer Expert, il maîtrise aussi bien le développement back-end (Python, C#) que le DevOps (Docker, Kubernetes, Terraform) et les agents LLM. Formateur en Python, cloud, DevOps et IA générative appliquée, il forme avec une obsession : Amener chaque apprenant à concevoir et déployer des architectures réellement scalables en production.

» Learn More