MLOps Community Amsterdam Meetup

Welcome to a new live get together for the global MLOps.community in Amsterdam. We'll have a combination of talks, lightning talks and ample time to socialize with others in the community!

IMPORTANT: We will need to provide the venue with your email to send you a QR code that will allow you access. This email will not be used for anything else.

Schedule:
17:30 — 🍕 Arrival
18:00 — Welcome
18:15 — 🎤 Model Deployment with KServe - Raahul Dutta
18:45 — 🎤 Machine Learning Quality at Booking.com - George Chouliaras
19:15 — 🥤 Networking
20:00 — Closing

Sign-up instructions:

• Sign up via meetup.com
• Let us know if you want to give a ⚡️ lightning talk (first come first serve)
• Let us know if you have any strict dietary restrictions (e.g. vegan🌱)
• We are looking for speakers for the next events. If you would like to give a talk, let us know the topic and a contact information.

**Machine Learning Quality at Booking.com**
George Chouliaras

Building successful Machine Learning Systems involves scientific fields like Statistics and Calculus as well as engineering fields such as Software Engineering. Only recently has the latter gained focus and relevance. At the same time, Machine Learning as a field poses new challenges to this well established discipline, Software Quality is one of them. In this talk we explain how Booking.com handles Machine Learning Excellence in a centralized way and we also present a novel Software Quality Model for Machine Learning which determines what quality in Machine Learning Systems is, and provides Machine Learning practitioners with a tool to holistically evaluate them.

Model Deployment with KServe
Raahul Dutta

Machine learning model deployment is not a painless job. There are considerable hardships around it like Cost, Model monitoring, Rollouts, Protocol Standard(Rest, Kafka, or gRPC), Batch prediction, Different type of Model integration(Pytorch , Tensorflow, Xgboost et), explainability, pre-post processing code handling etc etc. We , in Elsevier, created a KServe pipeline to manage the noted problems. KServe Provides Performant, Standardized inference protocol across ML Frameworks. It supports modern serverless inference workload with Autoscaling including a scale to zero on GPU, Simple and Pluggable Production Serving including Prediction, pre/post-processing, monitoring, and explainability including Advanced deployments with Canary rollout, transformers, experiments, ensembles, Model-Mesh.

🎤 The Speakers

• George Chouliaras

George Chouliaras is a Senior Machine Learning Scientist at Booking.com. He has more than 4 years of experience on building and deploying Machine Learning Systems at Booking.com, particularly in the area of Customer Service. George is one of the early advocates for improving the software quality of the Machine Learning Systems at Booking.com and is currently the tech lead of the ML excellence team, aiming to standardize and improve the ML development standards company wide. Prior to Booking.com, George obtained a Master’s Degree in Business Analytics from the VU university in Amsterdam, as well as a Diploma in Electrical and Computer Engineering from the Aristotle University of Thessaloniki in Greece.

• Raahul Dutta

Raahul Dutta is an MLOps Lead at Elsevier. he has around 6+ years of experience in transforming the Jupyter Notebooks into low-latency, highly scalable - production standard endpoints, he implemented various ML/AI models and pipelines (30+) and exposed them. He was associated before Oracle, UHG, Philips. He filed around 13 patents(some of them in `granted` status) in the ML, BMI and chatbot domain. He enjoys riding motorbikes and lives in Amsterdam with his partner.