Description

Training Overview

Duration: 1 Hour

Getting your data RAG-ready for usage within LLM’s is no easy task. When your data consists of various patient chart document types (e.g. PDFs, HTML, PPT not to mention handwritten notes) you will have your work cut out for you. Join us in a workshop where we will structure healthcare patient data for usage with LLM’s. With the help of ‘unstuctured.io’ packages we will transform, clean, chunk, generate summaries and embeddings, then write our ‘structured’ healthcare data to a vector database (weaviate) where it can be used to interact with LLMs or Chatbots.

Instructors Bio

Audrey Reznik Guidera

AI Platform Specialist Solution Architect | Red Hat

Audrey is an AI Platform Specialist Solution Architect working for Red Hat. She focuses on helping customers with managed services, AI/ML workloads and next-generation platforms. She holds a degree in Computer Information Systems and has been working in the IT Industry for over 20 years from full stack development to data science roles. Audrey is passionate about AI and in particular the current opportunities with AIML at the Edge and Open Source technologies.

Bob Kozdemba

Principal Specialist Solution Architect | Red Hat

Bob is a career pre-sales engineer with hands-on experience using best of breed open source technologies to solve AI/ML workflow challenges including MLOps, RAG and generative AI.

Training Outline

  • Why is Structuring Healthcare Data so important when creating a Healthcare LLM?

  • Discuss RAG Architecture for a future Healthcare LLM and/or Chatbot.

  • What does Unstructured.io do?

  • Why do we use a Vector Database?

  • Steps to structuring our data for usage!

  • Storing our structured data in Weaviate.

  • Querying our newly structured healthcare data.

Key Takeaways:

  • Structuring healthcare data (for Chatbot usage) is hard, but it is made easier by using data curation tools (such as Unstructured.io) and RAG.

  • Vector Databases are key to creating a useful LLM. The success of your LLM (or Chatbot) is directly related to the ‘curation’ of your data set.

  • Be smart about where (or if) you should move your data for curation and usage. Sometimes curating your data in place, then moving your vector database elsewhere is a good cost savings decision.

Background Knowledge:

  • Beginner knowledge of Kubernetes (or OpenShift), Python, Generative AI and database principals.

  • No medical healthcare data knowledge is necessary.