RAG and ROLL - Leveraging Python and Smart Retrieval with LLMs


Details
Join us for a technical-focused PyData event in Johannesburg. This meetup is ideal for Python and data enthusiasts seeking in-depth discussions without sales pitches. Connect with peers in a relaxed atmosphere and hear from expert speakers.
Schedule:
16:30 - 17:00 Welcome and Meet & Greet with Shaun de Ponte
17:00 - 17:20 Not your typical RAG application by Sheena O'Connell
- Rag apps are usually chat apps. The user enters some text, the application then does a vector search to retrieve some information from a database, then the data is augmented by an LLM that packages the information in a nice message.
This pattern is quite common because it is useful. But there are other ways to do things.
In this talk I'll show you how I used an open source vector search engine and some prompt chaining to solve a pretty niche problem . The problem itself isn't that interesting, the main thing I want to get across here is that the tools used to build RAG apps can be applied creatively, to solve different types of problems.
17:20 - 17:40 Building a Training Course Outline with Azure OpenAI by Tyrone Kruger
- Course planning doesn’t have to be complicated. In this talk, I’ll show you how I built an AI-powered bot that generates personalised course outlines based on career goals. The bot uses RAG to boost the accuracy of its recommendations. We’ll go beyond the usual use cases and explore how RAG tools can be used to simplify educational planning and rethink professional development.
17:40 - 18:00 Enhancing Retrieval Augmented Generation with GraphRAG by Luis de Sousa
- GraphRAG, developed by Microsoft Research, takes RAGs to the next level by moving beyond basic semantic search. Instead, it turns text into knowledge graphs and community hierarchies, providing a more nuanced context for generating content. In this session, I'll demonstrate how GraphRAG's unique approach boosts the accuracy and relevance of RAG outputs and how these same principles can be applied to solve various problems.
18:00 - 18:30 Refreshments and Networking
Theme: RAG and ROLL - Leveraging Python and Smart Retrieval with LLMs
Date: Tuesday, 10 September 2024
Venue: Microsoft Campus in Bryanston, South Africa, Auditorium 3
Location: https://maps.app.goo.gl/Typ7CfiYgCuRuntL7
Special Thanks to Microsoft South Africa for providing the venue.
Interested in Speaking? Submit your session abstracts at https://sessionize.com/pydata-johannesburg/
We look forward to seeing you there!

Sponsors
RAG and ROLL - Leveraging Python and Smart Retrieval with LLMs