As a Machine Learning Engineer at H&M you will join our content personalization and platform team that develop scalable and reusable components needed to enable real-time personalization use cases. These components cover model serving of supervised/unsupervised machine learning models, streaming of real-time events incl. storing & abstracting this data.
· Automatic model training and deployment
· Setup and automate elastic infrastructure in cloud
· Large scale A/B testing
· Etc.Required skills/qualifications:
· We are cloud native on Azure, so people with Azure experience will be beneficial.
· The online model serving will be based on Azure Kubernetes Service, on top of Kubernetes, most likely we will leverage https://www.seldon.io, we might use their commercial solution Seldon Deploy, so it will also include broad scope, like monitoring with Grafana, Prometheus, logging with Elastic/Azure log analytics, service mesh like Istio, GitOps with Argo CD for deployment version control
· We also need to build feature store for both historical customer profile and also online customer event. We do have some data engineer to cover that, but would be good that people has experience with feature store for online serving, we will probably leverage Azure Cosmo DB as look up table, also Azure Event Hub as message
· We are using Azure DevOps for CICD
· In additional, would be good to have at least one specialist with strong QA experience, draft test spec for stability test, performance test, e2e integration test etc.
· For InfraAsCode, would be good have at least one person, focus on Infra automation, also network, security setup etc. Then the person should have enough Azure experience.
· For model development, we love Databricks and Mlflow (for model management)Technical skillset
· Python > 5 yrs
· Software Development > 5 yrs
· (Py)Spark > 2 yrs
· Machine Learning > 2 yrs
· Data Wrangling > 3 yrs
· Model Management > 1 yrs
· Public Cloud > 1 yrs
· Bonus: DevOps
· Well recommended on teamwork, communication/presentation skill, and being creative
· Exposure to agile ways of working, e.g., scrum or kanban
Required language skills: