JoynJoyn
SHARE
More

Confidence

1

Pane of Glass

60+

Free Channels

Jun. 23, 2022

See how Joyn accidentally built a modern data stack powered by Prefect.

Industry

Communications

Company Size

200-499

Key Use Cases
  • Orchestration

Products Used
  • Prefect 1.0

  • Cloud 1.0

Have you ever turned on your TV during an hour of free time and relaxation and immediately felt overwhelmed or daunted by the number of choices you were about to make? Which streaming provider should I choose? Then which show should I watch? According to a 2022 study by Nielsen, 46% of viewers feel crippled by indecision due to the sheer volume of choices available to them, which makes what should be a relaxing moment into something stressful.

Joyn’s mission is to help you take the negative engineering out of finding the right streaming platform. Created as a joint venture between ProSiebenSat.1 and Discovery, Joyn aims to create a comprehensive German streaming platform of cross-provider content. Joyn already provides more than 60 TV channels and various free on demand content, multi-channel support on iOS and Android, and best of all, no subscription required. At a time when other streaming providers are beginning to feel more and more like the old days of cable, Joyn is dedicated to providing a truly modern and relaxing streaming experience.

Layered behind Joyn’s modern UI and innovative recommendation engine is a powerhouse data engineering team, leveraging a cutting edge modern data stack to fuel business decisions based on critical data insights. Modern data stacks are usually defined by their physical properties (cloud based, modular, flexibility), and each component must represent the best tool fit for purpose. Companies that focus too much on individual tools, however, lose sight of their main focus: in the case of Joyn, this was faster and easier delivery of insights from their data.

In service of this goal, Joyn chose Prefect because it was the most flexible, Pythonic tool in its space - the best in class when it comes to orchestration.

"We chose Prefect over other orchestration tools because it is easy to learn as you write workflows the same way as you would write standard Python apps,” - Director of Engineering, Simon Hawe

Joyn’s team was leveraging GCP (BigQuery), Snowflake, and Databricks to store and access their business critical data, and Great Expectations to ensure data quality. But if the modern data stack can be represented as a series of boxes and arrows, then these tools represent the boxes, and Joyn knew that they’d need a cutting edge orchestrator to manage their data pipelines between all of their tools.

"We were always looking for the right tool to fulfill the purpose. Following this approach, we accidentally ended up with an architecture that is coined as ‘the modern data stack’. This is something we are all proud of.” - Director of Engineering, Simon Hawe

“Prefect is easy to extend and integrate with other tools. Overall, this leads to a quick transition from PoC Code to production readiness and hence reduced time to produce business value.” Senior Procurement Manager, Angelo Malacarne

Prefect is designed to be minimally intrusive to the developer experience, while still providing the Joyn team the ability to design robust pipelines at scale.

“Our data team serves as the central hub for collecting and delivering high-quality data to the entire organization. We enable our product department to make data-driven decisions, as well as our BI department to compute accurate KPIs. Therefore, we all need to be confident that our pipelines are designed against failure, and in the rare case that they do fail, that we’ll be made aware. That is what Prefect delivers to us.” - Director of Engineering, Simon Hawe

With modern, intuitive orchestration powering Joyn’s modern data stack, the team is able to deliver not just data, but insights. In the last year, Prefect data flows have helped the Joyn data team to

  • increase confidence in business progress reports due to improved visibility and monitoring

  • provided a scalable process for data science team to conduct experiments

  • empowered UI/UX research through OOTB and custom integrations

all helping Joyn viewers to experience a more modern and intuitive streaming experience.

Posted on Jun 23, 2022
Case Study
Snowflake
Databricks
GCP
BigQuery
Great Expectations
Dynamic DAGs
Mapping
Error Handling
Logging
Scheduling
Integrations

Love your workflows again

Orchestrate your stack to gain confidence in your data