Past Meetup

PyData Amsterdam: Data Science week edition @ Flow Traders

This Meetup is past

105 people went

Location visible to members


It's PyData Amsterdam time!

First of all: the videos from the conference have been posted. Find them at the PyData Youtube channel (

As for the meetup: this edition takes place during the Dutch Data Science Week. You can find all about it here (, but long story short: if you're quick you can avoid cooking the whole week, as it's full with meetups and events!

The fine folks at Flow Traders will host us. The program looks as follow and as always, remember to use your real name when signing up!

• 18:00 Doors open

• 18:15 Dinner

• 19:00 A word from our host

• 19:05 Giovanni Lanzani: Data pipelines at scale with NiFi and Python

• 19:50 Break

• 20:00 Stephen Galsworthy and Kaustav Basu: Data Science with the Toon smart thermostat

• 20:45 Drinks and networking

• 22:30 Door closes (Yes, later than usual! We've heard they have extremely good alcoholic beverages, hence the extra hour).

First talk: Data pipelines at scale with NiFi and Python


Apache NiFi is a data flow tool that combines the ease of connecting built in data processors with the flexibility of creating your own processors using Java, Scala, Groovy, Python, etc.

In this tool we will introduce NiFi, show how you can create a simple pipeline to parse incoming files, route them based on their content, transform them using Jython, and finally send them to HDFS for long term storage.

We will go a bit deeper on how you should structure your Python application to achieve better performance.

We will also show how other sources and sinks can be used, such as Kafka, Cassandra, ElasticSearch.


He's that guy with lots of kids.

Second talk: Data Science with the Toon smart thermostat

Toon, one of Europe’s leading smart thermostats, is installed in over 350,000 Dutch homes. Toon combines a smart thermostat which controls the heating system together with an energy insight display and a smart home hub which connects to lighting and smoke alarms.

The Data Science team at Quby are currently developing numerous data driven services using Toon data. In this talk we will describe how energy data is collected from Toon and processed using Python and Spark. We will explain how we apply disaggregation algorithms to give our users insights to optimize their gas and electricity usage.

If you’d like the opportunity to work with Toon data during Dutch Data Science Week then please sign up for our two day Eneco Smart Home hackathon (!


Stephen Galsworthy is the Head of Data Science at Quby, a leading European producer of smart thermostats. In this role, he is responsible for the development of data driven services for residential energy customers and utilities.

Before this, he worked as a consultant at TNO developing smart grid technology for the Dutch and European markets. Prior to this, he led a Data Science and Engineering team in London focusing on developing novel demand response technologies. Stephen holds a Ph.D. in Mathematics from Oxford University.

Kaustav Basu is a Data Scientist at Quby currently engaged in researching, prototyping and deploying data driven services. Previously, he worked both in the academia and industry. He holds a Ph.D. in Computer Science from University Grenoble Alps, France.

How to get to Flow Traders

Public transport:

• Take tram 26 from the Central Station, direction IJburg. Step off at the Rietlandpark stop and use the escalators to get on the street level. Cross the road direction the train underpass. When you emerge from the underpass, turn right direction the large parking lot. You will see the large office building (INIT). The main entrance is on the other side of the building.

• Take tram 10 from the city center, direction Azartplein. Get out at 1e Leeghwaterstraat. Turn left to Tweede Leeghwaterstraat, then right to Conradstraat. After ca 50 m, enter the large parking lot. The office building is right in front of you. The main entrance is on the left side of the building.

By car: Ring Amsterdam A10, take the exit 14, Zeeburg. Navigate to Jacob Bontiusplaats 9. There is sufficient space to park your car in front of the INIT building (paid).