Agentic AI: An Exploratory And Functional Approach documentation
This documentation presents the design, architecture, and implementation details of a modular agent-based system developed for the final degree project (TFG). The system combines large language models with external tools for intelligent decision-making in scientific and environmental contexts.
Getting Started
Project Structure
The system is structured into the following modules:
Agents: Specialized LLM-based agents for handling specific tasks.
Tools: Interfaces for external APIs and databases.
System: LangGraph pipeline for coordinating agentic reasoning.
Evaluation: Scripts for automatic performance testing.
Utils: Configuration and helper utilities.