Build smarter LLMs with dynamic data retrieval

🔍 Workshop overview

Did you know that almost all modern Large Language Model (LLM) applications have Retrieval-Augmented Generation (RAG) as a core component? This hands-on workshop is designed for developers and ML engineers to build powerful LLM applications leveraging RAG, and explore techniques to enhance them beyond basic proof-of-concepts.

Participants will gain a solid understanding of core RAG architecture and learn strategies to improve the retrieval mechanisms that feed into LLMs, enabling them to handle complex queries with advanced methods and AI agents. Through a mix of foundational theory and practical exercises, this workshop bridges the gap between simple prototypes and production-ready systems.

By the end of this workshop, participants will be equipped to develop, refine, and optimize RAG-powered LLM applications ready for real-world challenges.

Note

• This workshop is of 4 hours duration.
• It is an in-person and hands-on workshop.
• Basic knowledge of Python and experience with OpenAI API is recommended.
• Participants must bring their own laptops and OpenAI API keys.
• Code & materials:
  • 📖 Slides
  • 💾 Exercises

🧭 Agenda

Segment 1: Foundations of Retrieval-Augmented Generation (RAG)

Duration: 2 hours (including breaks)

• Poll: Pulse check on participants’ understanding of RAG

• Foundations of RAG (30 min)
  Introduction to RAG architecture: semantic search, embeddings, and vector stores
  Simple demo for embeddings

• “Naive” RAG Architecture (30 min)
  Understand a basic RAG setup
  Discuss the interplay between retrieval mechanisms and generative models

• Q&A

• Building a Basic RAG Example (30 min)
  Hands-on: Create a simple RAG application using Python

• Challenges with “Naive” RAG Implementations (20 min)
  Discuss limitations and impact on application performance

• Q&A

Break: 10 minutes

Segment 2: from PoC to Production - enhancing RAG responses

Duration: 1 hour 45 mins (including breaks)

• Techniques to Improve “Naive” RAG performance (30 min)
  Advanced optimization strategies:

  • Reranking
  • Hybrid search
  • Metadata filters

  Q&A

• Hands-on: Refined RAG Implementation (30 min)
  Refine earlier implementation with advanced techniques
  Evaluate and showcase improved performance

• AI Agents and Orchestrating Complex Workflows (25 min)
  Learn how AI agents can improve reliability and performance
  Examples: Self-RAG, Corrective RAG

• Q&A

💻 Prerequisites

• Basic knowledge of Python
• OpenAI API key and experience with OpenAI API
• Familiarity with tools like ChatGPT
• Laptop with development environment set up

👥 Who should attend

• Developers with a basic understanding of Generative AI
• Engineers working on proof-of-concept RAG applications and aiming to make them production-ready
• ML/AI Engineers looking to deepen their applied knowledge of RAG systems
• Developers and technologists interested in becoming AI Engineers

📚 What will participants learn?

• Understand RAG architecture and its role in modern LLM applications
• Build a basic RAG-powered LLM application
• Learn common limitations of naïve RAG and practical ways to address them
• Apply advanced techniques such as reranking, hybrid search, and metadata filtering
• Explore how AI agents can orchestrate workflows and improve LLM application performance

👨 🏫 Instructor bio

Sarang Sanjay Kulkarni leads a healthcare client account at Thoughtworks, delivering generative AI-powered research assistants to expedite drug discovery. With over 13 years of experience across development, DevOps, data engineering, and AI engineering, Sarang brings deep technical expertise to his workshops.

He is an experienced trainer, having conducted numerous developer bootcamps, data engineering programs, and generative AI training sessions. Sarang is also a trainer at O’Reilly, where he conducts a workshop on building RAG-based LLM applications.

How to attend this workshop

This workshop is open for The Fifth Elephant members and for The Fifth Elephant 2025 annual conference ticket buyers.

This workshop is open to 30 participants only. Seats will be available on a first-come-first-serve basis. 🎟️

Contact information ☎️

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