Thu Feb 23 2023

Integrating an ML Library into Software Stack & Cloud Analytical Databases

Raamstraat 7 · Amsterdam, NH

Details

Data & Drinks event returns on Thursday February 23 at 17:30 at Xomnia's HQ. Our guest speakers will discuss two topics: Integrating a machine library into software stack & cloud analytical databases.

Our first guest speaker is Machine Learning Developer Shady El Gewily, who will discuss the topic of developing a machine-learning library for an existing software product by taking the road less traveled.

Our second speaker is Data Architect/ Data and Analytics Engineer Rogier Werschkull. He will dive into cloud analytical databases, discussing which of those databases to choose and why.

The event will include dinner, drinks and a lot of networking opportunities with data professionals from Amsterdam and beyond.

Short summary of the talks:

  • Talk #1 by Shady El Gewily: The Slimstock research team was founded to integrate machine learning into Slim4, the Java-based web-application that helps 1300 customers worldwide to optimize their inventory. Slimstock developed an ML library, which is used as a toolbox for the development of ML applications that are useful for inventory optimization. This ML toolbox is embedded in Slimstock's software, interfaced to their customers; data and running on their hardware. Due to the way our company is structured and our software is deployed and used, we made some unconventional choices for our stack.

In this talk, Shady will reflect on the challenges that our unconventional approach presented, ranging from model training speed, incorporating automated testing and ensuring models are backward-compatible. Knowing what we know now, what are better choices for the future?

  • Talk #2 by Rogier Werschkull: Analytical databases are specialized databases optimized for analytical use cases. The primary advantage of this type of database compared to OLTP databases is scalable aggregation queries on large amounts of data. Which analytical databases are available in the public cloud? How do they compare and in what aspects do they differ? Which database do I need to choose for my client based on which criteria? In this session, Rogier will attempt to answer those questions.

Rogier will also compare Snowflake, Bigquery, Redshift and Azure SQL pools (previously Azure SQL data warehouse) based on costs, scalability and developer / user criteria.

Get to know our speakers:

  • Shady el Gewily is a machine learning developer at Slimstock, where he is part of a team that develops a library for building ML applications that are useful for inventory optimization. Previously, he worked as a freelancer and developed a solution to automate the optimization of email marketing campaigns subject to constraints He is passionate about the entire ML product lifecycle, from prototype to model monitoring.
  • Rogier Werschkull has worked in data and analytics since 1999. Throughout his career, he has had a diverse set of roles, such as BI architect and data warehouse architect. In his current role, Rogier works as a freelance data architect and data / analytics engineer at the health data startup '6 gorillas' in Utrecht.