Nb. days
3
Face-to-face?
Yes
Remote?
Yes
Min. participants
2
Max. participants
8
Place
Toulouse, Paris, etc.
Pricing
2290€ HT
Artificial Intelligence is undergoing a profound transformation. The current enthusiasm for AI agents and Agentic AI stems from their ability to overcome the main limitation of previous technologies: the lack of proactive autonomy for complex tasks. Agents are transforming AI from a simple reactive tool (that awaits instructions) into an autonomous collaborator (that can plan, reason, and act) to achieve complex objectives without constant human intervention.
This intensive 3-day training is designed for all individuals with technical data skills who are ready to evolve their LLM applications from a linear stage to a dynamic and intelligent application.
This intensive 3-day training is designed for all individuals with technical data skills who are ready to evolve their LLM applications from a linear stage to a dynamic and intelligent application.
At the end of this training, you will be able to design systems capable of executing complex tasks in multiple steps and adapting to their environment using LangGraph, the most powerful framework for creating sophisticated reasoning loops and multi-agent systems.
Goals
- Understand the Philosophy: Master the fundamental principles and strategic value of Agentic Artificial Intelligence (Agentic AI) – planning, reflection, action.
- Master the Technical Foundations of LangChain.
- Implement Autonomy: Design and build autonomous agents based on the ReAct reasoning model with LangGraph.
- Orchestrate Collaboration: Model and develop complex systems involving multiple specialized agents and a Router Agent.
- Optimize and Deploy: Apply cost optimization techniques, ensure agent quality and security, and prepare for their deployment in a production environment.
To know the dates of the next training sessions,
to adapt this program to your needs
or to obtain additional information,
contact us!
to adapt this program to your needs
or to obtain additional information,
contact us!
Teaching method
- This training is 70% practical and uses illustrated and didactic exercises.
- A daily evaluation of the acquisition of the knowledge of the day before is carried out.
- A synthesis is proposed at the end of the training.
- An evaluation will be proposed to the trainee at the end of the course.
- A course support will be given to each participant including slides on the theory and exercises.
- A sign-in sheet for each half-day of attendance is provided at the end of the course as well as a certificate of completion if the trainee has attended the entire session.
- A follow-up and an exchange with the participants will take place a few days after the training.
Prerequisite
- Knowledge of the Python language is necessary.
- Basics in machine learning, LLM, and RAG are required, or having completed the Mastering LLMs and RAG for Generative AI training.
Targets
- Data Engineers
- Data Scientists
- Developers
Evaluation
This training does not require a formal assessment of learningProgram
Agentic AI: Theory and Challenges
- Definition of Agentic AI: the Perception, Reasoning, Action, Learning cycle.
- Key Differences: Traditional Agents vs. Generative AI vs. Agentic AI.
- The ReAct (Reasoning and Acting) reasoning model and its importance.
- Role of Tools and Prompts for reliable reasoning.
Designing an autonomous agent with LangChain
- Foundations : LLMs and Prompts, Tools (Outils), Retrievers and Vector Stores :
- Assembly : LangChain Expression Language (LCEL): Concepts of Runnables, Chains
- Transition to LangGraph
Discovering the LangGraph Framework
- Introduction to LangGraph: why orchestrate agents?
- Fundamental concepts: Graph, State (shared state), Nodes, and Edges.
The Autonomous Agent Triptych: State, Tools, and ReAct Engine
- Creating a structured State for the graph (input/output management)
- Definition of Python functions (@tool)
- Implementation of the ReAct loop
Agent Improvement and Robustness
- Advanced State Management and Memory
- Self-correction and resilience
Multi-Agent Orchestration
- Discovery of Multi-Agent Architectures
- Multi-Agent Interaction Models: Pipeline vs Group Chat
- Management of flows with the Router Agent
- Theoretical Concepts of Agentic Communication: Introduction to emerging A2A and MCP protocols.
Complex Systems, Optimization
- Modeling & Implementation of a Workflow in chain
- Management of approval and consolidation cycles
- Cost and performance optimization
- Implementation of barriers
Quality, Deployment & Perspectives
- Agent Quality: Observability, Logging, Tracing, Evaluation, Metrics
- From prototype to production: Good practices & use of A2A
- Future trends and interoperability
