Big data architecture - choosing the right tech for your app

Target audience - SW and data engineers

Background
Riskified is an AI platform used by online merchants to maximize eCommerce revenue. A big part of this is done by analysing the browsing session of the shoppers.

What We'll Be Doing:
18:00-18:30 : Cold beers, snacks, meet new people
18:30-19:15: Session #1: Drorit Baron, Software Developer, Riskified
19:15-19:30: Break for more beer + pizza!
19:30-20:15: Session #2: Inbar Yogev, R&D Team Leader, Riskified

*Sessions will be in Hebrew

Session #1: Umm scratch that: A tale of big data architecture redesign - Drorit Baron, Software Developer, Riskified

Riskified is an AI platform used by online merchants to maximise the shopping experience on their site. A big part of this is done by analysing the browsing session of the shoppers.
In this talk, Drorit will describe in details how we as Riskified designed and executed a full rewrite of our clickstream data pipeline.
She will present how we replaced our whole technology stack, moved from batch processing to stream processing, reduced end to end latency and response time, increased throughput and built a highly scalable system, all while significantly reducing our cost.

What you will learn:

• Big data processing methods and best practices.
• What considerations are important when choosing the best technologies and the right data model for your system as your company grows.
• We will shortly discuss whether a full rewrite is something that can or should be avoided.

Session #2: Data Modelling bling bling - Inbar Yogev, Data Engineering Team Leader, Riskified

Data modelling done right is in many cases the key for life without scaling pains.
A poorly designed/naive data model can often become a real pain as the business grows, and usually results in:

• More money being thrown on resources
• More time spent on development hacks & maintenance

What you will learn:

• How to model data for scale given a real-life use case.
• How different database technologies can do the trick given the "correct" data model.

Presenters:
Drorit Baron - https://www.linkedin.com/in/drorit-baron-b0182979/
Inbar Yogev - https://www.linkedin.com/in/inbar-yogev-981b5161/

_______
Parking (paid): Mohiliver St 18 (Central Park)
Grozenberg St 16 (Central Park)
HaCarmel parking (אחוזת חוף Kalisher St corner of Carmel St)
Bus Station: Carmelit Terminal, Allenby Maze Station