Senior ML Engineer Recommender Systems, Developer Advocacy | Germany | Remote
Grafana Labs is building an Interactive Learning system, an open-source, in-product learning experience that helps users succeed within the platform. A central part of that vision is a personalized recommendation system that helps each user discover the next guide, action, or product experience most likely to help them succeed. We are hiring an ML Engineer to lead its evolution into an increasingly personalized, continuously improving system driven by real-time product behavior, content metadata, customer context, and experimentation.
What You’ll Be Doing:
- Evolve Pathfinder’s recommendation system: Develop personalized approaches to candidate selection, ranking, sequencing, and next-best-action recommendations, while owning a real-time recommendation service.
- Build and operate applied models: Develop, validate, version, monitor, and iterate on models used by the recommendation system, owning the end-to-end training and serving.
- Define recommendation quality: Develop offline, online, and longitudinal measures of performance and own feature pipelines and monitoring of the model architecture.
- Ship incremental improvements: Use existing infrastructure to integrate improvements iteratively rather than waiting for a complete replacement system.
- Partner across disciplines: Work closely with software engineers to productionize models, collaborate with the Product Analytics team on metrics and instrumentation, and translate ambiguous needs into testable hypotheses for Developer Advocacy and Product teams.
What Makes You a Great Fit:
- Recommendation and personalization science: Proven experience building recommendation, ranking, search, matching, or next-best-action systems, utilizing simple, explainable approaches where appropriate.
- Distributed Systems: Practical experience with HTTP/gRPC, streaming technologies, and programming in Go or TypeScript.
- Applied model ownership: Experience personally building, validating, and monitoring models in production environments, with the ability to work effectively in version-controlled codebases.