Build multi-agent workflows with LangGraph

What we’ll do

The rise of agentic workflows marks a significant shift in how we harness the capabilities of Large Language Models (LLMs) to perform complex, multi step reasoning and decision making tasks. Unlike traditional pipelines, where prompts are static and behavior is tightly scripted, agent based systems offer dynamic orchestration. LLMs can choose tools, recall memory, decompose problems, and iterate toward a goal.

This hands-on workshop is designed for experienced developers, data scientists, and graduates eager to build real-world agent workflows using LangGraph, a cutting-edge framework for multi-agent orchestration.

What you’ll take away

• Foundations of LangGraph: understand nodes, edges, graph state management, and type-safe execution flows.
• Designing RAG (Retrieval-Augmented Generation) workflows: build hybrid systems combining retrieval and generation using LangGraph constructs.
• Persistence, memory, and Human-in-the-Loop (HITL): learn how to persist agent state, inject memory, and bring human decision-making into the loop.
• Capstone project: apply your knowledge to build a complete end-to-end AI system by the end of the day.
• Advanced LangGraph patterns: dive into sequential, branching, and parallel workflows. Explore evaluator-optimiser and multi-agent architectures.
• MCP & Agent-to-Agent (A2A) protocols: enable inter-agent communication and collaborative problem-solving using Model Context Protocols (MCP).

This workshop is deeply interactive and code-focused – expect to roll up your sleeves and build! By the end of the day, you’ll walk away with a working multi-agent prototype and a clear understanding of how to architect complex AI systems.

Who should attend

• Senior software engineers and data scientists
• ML/AI practitioners working with agents, LLMs, or LangChain
• Advanced graduates with strong Python & ML fundamentals

Workshop Prerequisites

To ensure participants can make the most of this workshop, please note the following prerequisites:

• Python proficiency: participants should have at least a basic to intermediate understanding of Python. The workshop will not cover Python fundamentals.

• LangChain experience:

  • Attendees are expected to have prior experience with LangChain. We will not revisit LangChain basics, so familiarity with its core concepts and components is essential.

• API key management:

• The workshop will not provide any API keys. Participants have to obtain and manage their own API keys (e.g., OpenAI, Table Research, LangSmith, etc.).
• Set up working API keys in advance, as we will use them for hands-on exercises (including tracing, model calls, etc.).

• Development environment:

• Participants should have a functional Python development environment (e.g., Anaconda, venv, or similar) set up on their laptops.
• Familiarity with installing packages via uv or pip or conda is expected.

• Familiarity with Jupyter Notebooks or IDEs:

• Exercises and examples may be provided in notebook or script format. Please be comfortable running code in your preferred environment.

• General AI Concepts (recommended):

• While not mandatory, a general understanding of key AI concepts (such as prompt engineering, LLMs, and vector databases) will help you get more from the workshop.

About the instructor

Suman Paul Choudhury is a Senior Data Scientist at Sahaj with over 9 years of industry experience in advanced data science and AI. He holds a Master’s degree in Signal Processing and specialises in computer vision, NLP, generative AI, and deep learning. Throughout his career, he has developed and deployed a wide range of AI solutions tailored to diverse client needs, driving measurable business impact.

Duration

This is an in-person workshop of 8 hours duration.

How to attend this workshop

This workshop is open for participation to The Fifth Elephant annual members
This workshop is open to 40 participants only. Seats will be available on first-come-first-serve basis. 🎟️

Contact information ☎️

For inquiries about the workshop, contact +91-7676332020 or write to info@hasgeek.com