Machine Learning 
Every listener finds it easier and more fun to choose what to play next thanks to the Personalization team.  We are responsible for some of Spotify's most beloved features, like AI DJ and Discover Weekly.  We created them by having the most in-depth knowledge of the podcast and music industries. 
If you join us, you will continue to keep millions of users listening by giving each and every one of them excellent recommendations and insightful context. 
Location 
Stockholm 
Job type
Everlasting  In order to please listeners, would you like to assist Spotify in creating new tailored sessions using generative voice AI?  In this position, you will collaborate with Speak, Spotify's Text-to-Speech (TTS) team, to provide voice-generated audio that enhances users' listening experiences and podcast suggestions. 
What You'll Do
Work with a heterogeneous team to make sure machine learning models are scalable and extremely effective for production use cases.
Create and construct effective machine learning model serving infrastructure that enables extensive deployments in many locations.
For real-time serving and production applications, optimize machine learning models in Pytorch or other libraries.
Work closely with academics and machine learning engineers to spearhead the process of bringing machine learning models from research and development into production.
Create and manage scalable Kubernetes clusters to deploy and manage machine learning models while guaranteeing performance and dependability.
To ensure production stability, implement and track logging metrics, identify infrastructure problems, and participate in an on-call schedule.
Influence decisions about infrastructure and technical design to support novel and varied machine learning architectures.
Work together with interested parties to advance projects pertaining to the large-scale serving and optimization of machine learning models.
Who You Are
You are passionate about generative machine learning, audio, and/or speech.
You have a great deal of experience with machine learning frameworks like Pytorch and are an expert at optimizing machine learning models for production use cases.
You have expertise operating Kubernetes clusters in multi-region configurations and creating effective, scalable infrastructure to support machine learning models.
You are at ease working with novel, cutting-edge architectures and possess a solid understanding of how to take machine learning models from research to production.
You understand how to diagnose production problems and write logging metrics, and you're willing to participate in an on-call schedule to preserve performance and uptime.
In order to develop and enhance model deployment pipelines, you like collaborating closely with research scientists, machine learning engineers, and backend engineers.
You do best in settings that call for resolving intricate infrastructure issues, such as performance optimization and scaling.
It is advantageous to have prior knowledge of low-level machine learning libraries (such as Triton and CUDA) and speed optimization for unique components.
Where You'll Be:
We give you the freedom to work where you are most productive!  As long as we have a work site, you can be anywhere in the European Union for this position.
For collaborative purposes, this team works in the GMT/CET time zone.
France is not included because of on-call limitations.
Our worldwide advantages:
plenty of educational options thanks to our hardworking staff at GreenHouse.
You can pick how you want to participate in our success with flexible share incentives.
All new parents are entitled to six months of fully paid global parental leave.
Our self-care center and staff support program is called All The Feels.
Public holidays are flexible; you can switch off days based on your ideas and ideals.
Discover what life is like at Spotify.
When Spotify first began in 2008, it completely changed how people listen to music.  By enabling a million creative artists to make a living from their work and providing billions of fans with the ability to appreciate and be fervent about these creators, we hope to unleash the creative potential of humanity.  We are motivated by our passion for podcasting and music in everything we do.  With a user base of over 500 million, we are currently the most well-known audio streaming subscription service worldwide.
Find out more about our culture.