MLOps Belgium meetup May 2024 (Gent)

May 2024 edition of MLOps Belgium meetup!

Event will be at the Lighthouse offices. A link will be shared so that you can follow remotely.

Link: TBD

Food and drinks will be available between 18h and 19h.

Also, don't register if you are not able to attend physically. A link will be provided to follow remotely but we want to use the attendance list to estimate the number of people being present for food and drinks. Please mind the organisational effort from the people involved. Thanks!

Overview of the talks:

MLOps with a feature store - a unified architecture for batch, real-time, and LLMs

Speaker: Rik Van Bruggen, Hopsworks

An important challenge in building ML systems is deciding how to structure you ML system - whether it is a batch ML system or a real-time ML system or whether you are fine-tuning a LLM, and then using RAG to connect it to external data.In this talk, we will present a unified architecture for these ML systems built around the idea of feature, training, and inference pipelines connected by a feature store. This architecture presents a clear set of abstractions with which to build different classes of ML systems - from batch to real-time to LLM systems. It is also lifts discussion of MLOps from technologies to principles, such as versioning, testing, and monitoring of ML systems.

Bridging the Gap: Empowering Developers with MLOps for Seamless Production Integration

Speaker: Jose Eduardo Mendoza Blanco, Prophecy Labs

In the rush to innovate and deploy machine learning models, the divide between development and production is often overlooked, leading to inefficiencies, delays, and even failures in deployment. This presentation confronts this neglected gap head-on, highlighting its impact on project timelines and collaboration Through real-world examples and case studies, we unveil how leveraging MLOps starting at the experimentation stage, we can empower developers with tools and processes to iterate swiftly, enhance collaboration, ensure reproducibility, and maintain model quality in production environments.