Data Science Meetup #14

Data Science Meetup Köln
Data Science Meetup Köln
Public group
Location image of event venue


Hey folks,

We are having our next meetup next week! Sign up soon! We have one speaker so far, but will surely find another one. So, stay tuned to get all the updates.

See ya soon!

PS: Data Science meetups will now take place on a regular 4-weeks basis! So save the date for March, 21st!

Topic 1:
AI meets CX
Christopher Harms, Lead Data Scientist at SKOPOS

Customer experience research aims to find the key drivers of why customers are happy – or unhappy – with your products and services. In times when surveys need to be shorter and open-ended questions are on the rise, gaining valuable insight in a scalable way is a challenge faced by the whole insights industry. In this case study we show how an elegant combination of NLP techniques, machine learning methods and statistical tools can provide actionable insights and helps SEAT to identify how to ensure customer satisfaction and loyalty. In this talk, we show how data science can provide scalable, valuable solutions to real-world business problems when time and resources are scarce.

Christopher Harms is Lead Data Scientist at the market research agency SKOPOS. Together with his team, he develops new tools and solutions to handling and analyzing data. He has a background in psychology and is still working on his PhD on statistics and replicability in psychological research.

Topic 2:
What is consensus and how to achieve it
Kaveh Vahedipour, Senior Architect at ArangoDB

Modern data stores must be scalable to 100s of storage nodes. While large server farms bring resilience through replication and performance through load balancing, it becomes a non-trivial task to establish a ground truth, configuration management, one time initialisation and other race conditions. Computer science has for decades been in search of dependable protocols which have mathematically been proven correct, that deal with consensus. Recent developments all rotate around the RAFT protocol. No matter how clear and correct a protocol is in definition, its realisation on a multi-user, time-sharing operating-system proves exceedingly hard to get right. At ArangoDB, we invested considerable amount of effort in such an implementation. This talk clarifies the notion of consensus from a data-scientific point-of-view and details its realisation in theory and practice.

Kaveh ist Senior Architect at ArangoDB and has a rich research and praxis background in Data Science. ArangoDB is a native multi-model NoSQL database