PyData Cluj-Napoca: Meetup #23

🎄 Celebrate the Year with Our Final PyData Meetup Before Christmas! 🎄

🌟 Hello PyData Community! 🌟

Join us for a festive evening as we wrap up the year with our last PyData meetup before the holidays. This special gathering will feature insightful talks on the latest developments in Python, ML and data science.

Whether you're a seasoned professional or just starting your data journey, this meetup is the perfect chance to reflect on the year's accomplishments, share your experiences, and get inspired for the year ahead.

Don't forget to bring good vibes and warmth for a social and informative evening! (+🍕🍔🌯🌮)

----------------------------------------------------------------------------------------------------
"Watt's Next? Using Python to Analyze and Cluster Energy Consumption Data" by Ioana Barboș

In a landscape of ever-increasing electricity consumption, overloaded electricity distribution networks and renewable energy funding at an all-time high, knowing how to work with time series and sequential data can be a data scientist’s competitive advantage.

We will scratch the surface of this field during this evening’s workshop, by using energy consumption data collected from smart meters installed on buildings in Portugal. In the process of getting to know the data, we’ll discuss different modelling approaches and explore ways to group our consumers based on their energy consumption behavior... all in Python.

----------------------------------------------------------------------------------------------------
"A Practical Introduction to RAGs" by Vlad Griguta

Retrieval-Augmented Generation (RAG) systems have recently emerged as a significant branch of NLP, combining semantic retrieval with LLM generation to create powerful knowledge extraction frameworks.

While RAG systems have proven effective, implementing them in enterprise environments presents unique challenges that go beyond basic setups. In this talk, we will explore these critical components and the obstacles that arise in elevating their performance, including:
1. Establishing a consistent evaluation framework
2. Enhancing semantic approaches with word-level retrieval
3. Optimizing the data ingestion flow for streamlined performance

We will begin with a standard RAG pipeline and progressively build complexity as we examine the limitations of each component. By the end of this talk, you will have the tools and knowledge to leverage this pipeline and customize your own RAG system.

----------------------------------------------------------------------------------------------------
NumFOCUS Code of Conduct
https://numfocus.org/code-of-conduct