A step-by-step guide to building multi-agent frameworks in the era of generative ai
This publication was co-authored with Rafael Guedes.
The rise of generative ai is the new platform shift of the digital age. It solves problems ranging from automation in large companies to various types of R&D and creativity. The global market is expected to exceed $65 billion by 2024, and 86% of IT leaders anticipate major organizational changes (1). So far, the biggest returns have come from chatbots (the most generic and abundant use case), code copilots, and enterprise search.
Investment continues to flow into ai, with $13.8 billion invested in 2024 (a six-fold increase from 2023) (1). Additionally, companies are incorporating ai into their core strategies and systems. Technologies such as recovery augmented generation (RAG), fine-tuning, and specialized models for vertical applications (e.g., healthcare, legal) are becoming widespread.
Large language models (LLMs) have brought attention to ai (in several ways) and opened the door to new ways of solving old problems. This new way is through Agentic ai, a framework in which autonomous agents work collaboratively to execute complex multi-step workflows.
Our demo shows how you can work and develop a multi-agent system. It integrates three specialized agents:
- TO web investigator agent…
