Your LLM-powered guide to the posit::conf(2025) agenda

posit::conf(2025) is just around the corner, and we can’t wait for an unforgettable three days in Atlanta, Georgia, from September 16th to 18th. Whether you’re a seasoned R or Python pro, a budding data scientist, or just curious about the latest innovations, there is something on the agenda for you.

Speaking of which, the full agenda for posit::conf(2025) is now available! It is packed with engaging workshops, insightful talks, and intriguing keynotes covering everything from foundational R and Python concepts to data visualization, integrated development environments, and responsible AI.

Start exploring the official posit::conf(2025) schedule to plan your personalized conference experience. However, we know that with so much to see and do, navigating the extensive list can sometimes feel overwhelming.

 

Introducing Sam Parmar’s LLM-Powered Shiny App: Your Agenda Co-Pilot

 

That’s why we’re thrilled to highlight a fantastic community-driven project that will change how you interact with the posit::conf(2025) agenda! Our very own community member, Sam Parmar from Pfizer, has developed a Shiny application that uses the power of large language models (LLMs) to help attendees explore the conference schedule. We have dubbed it the “posit::conf(2025) Agenda Chat Bot”.

CHECK OUT THE APP HERE

 

Use natural language to ask something like, “What are all the sessions related to Positron?”

 

Sam’s project is a great showcase of the power of open source tools for web scraping, data cleaning, LLM integration, app development, publishing, and more!

  • Sam used packages like httr2 and chromote to automatically pull the agenda details from the conf event portal.
  • Once he had the raw data, he cleaned it up using essential packages from the tidyverse (like purrr) and beyond.
  • To prepare the data for the LLM, Sam used the ragnar package to create a knowledge store that the LLM could search and retrieve information from. This is implemented with Retrieval-Augmented Generation (RAG).
  • He then built a user-friendly app with Shiny, connected it to an OpenAI model with ellmer, and created the chatbot interface with shinychat. When you ask a question, the LLM finds the answer in the knowledge store.
  • Finally, he published the app on Posit Connect Cloud, which securely stores the OpenAI API key so that others can chat with the model.

It is truly a testament to the incredible creativity within our community. He’s also shared all his code on GitHub so you can see how he did it yourself. Thank you, Sam!

We can’t wait to see you at posit::conf(2025)! Get ready to explore, learn, and connect like never before, with a little help from our amazing community and the innovative spirit of data science.

REGISTER NOW