Prerequisite: See below

Nicolas Decavel-Bueff

Nicolas Decavel-Bueff is a senior AI engineer at Candidly, specializing in deploying ethical AI solutions with a focus on large language models (LLMs). Drawing on a mix of open- and closed-source LLMs, he combines metric-driven evaluation with advanced retrieval and agentic techniques. He previously built ChatHue, Behr Paint's customer-facing multimodal chatbot handling roughly 10,000 conversations per month. Holding a master's in data science from the University of San Francisco, Nicolas empowers users in highly regulated industries to harness AI responsibly.

Kristy Hollingshead, Ph.D.

Dr. Kristy Hollingshead is an associate director of AI engineering at Further and a TDWI research fellow. She specializes in natural language processing (NLP), machine learning (ML), and the implementation of governance frameworks for safely operationalizing advanced AI technologies. With a Ph.D. in computer science from Oregon Health & Science University, she focuses on reliable, secure AI systems for highly regulated sectors, including healthcare, finance, and defense. Throughout her work, Dr. Hollingshead champions AI as an augmentative tool that empowers people, rather than replacing them.

Large language models (LLMs) have transformed the way organizations interact with their data through natural language. While these models can deliver impressive results right out of the box, their true business potential is unlocked when they are customized to meet specific needs. This hands-on workshop equips your team with the practical skills to implement retrieval-augmented generation (RAG) systems, empowering your LLMs to access, interpret, and leverage your organization’s unique data assets.

In this workshop, your team will learn to build and optimize a complete RAG pipeline—from selecting the right models and preparing data to applying effective retrieval strategies and evaluation techniques, while also implementing robust guardrails for responsible AI use. Through interactive labs and real-world case studies, students will explore both foundational concepts and advanced optimizations that make LLMs indispensable in business environments.

By the end of the session, you’ll gain hands-on experience with the essential components needed to develop, evaluate, and deploy customized LLM applications that generate measurable business value.

You Will Learn

• How LLMs function and the practical considerations in model selection
• The fundamentals of prompt engineering to enhance LLM performance
• How to build a complete retrieval-augmented generation (RAG) pipeline
• Techniques for evaluating and measuring RAG system performance
• Methods for identifying and resolving common RAG failure points
• How to implement tracing and guardrails for secure, responsible AI systems
• Best practices for scoping and executing RAG projects that deliver business value

Geared To

• Data scientists and engineers deploying LLM applications
• AI/ML practitioners seeking to enhance their RAG implementation skills
• Software developers integrating LLMs into business solutions
• Technical leaders planning LLM implementation strategies
• Business analysts exploring practical applications of generative AI
• Anyone ready to move beyond basic LLM usage to create tailored, business-specific AI solutions

Pre-requisites

Workshop exercises will feature pre-written Python code, meant to be run in an online notebook.

The instructor will guide you on how to run the code, but comfort reading code will help.

Laptop Setup

Attendees will need a laptop computer with access to specific web services. In advance of the class, attendees will receive detailed instructions for preparation.