Senior Machine Learning Engineer, Grafana Labs

• We are looking for someone who can understand the whole problem, establish strong foundations, and ship measurable improvements into the existing recommender one iteration at a time
• We know it is rare to find everything. Strong candidates should demonstrate credible ability across all three core areas below and be particularly strong in at least two
• Recommendation and personalization science: you have built recommendation, ranking, search, matching, propensity, or next-best-action systems. You are comfortable beginning with simple, explainable approaches when they are the best way to learn
• HTTP/gRPC, streaming, Go/TypeScript previous experience in distributed systems
• Applied model ownership. You have personally built, validated, monitored, and iterated on models used in a product or operational environment. You can work effectively in version-controlled codebases and collaborate with engineers on production implementation
• You should also be a strong product thinker and technical communicator
• You can take an ambitious and ambiguous objective, identify the most important unknowns, and create a sequence of models and experiments that steadily improves the product

Desirable

• Grafana Labs is building an Interactive Learning system, an open source, in-product learning experience that helps users learn and succeed without leaving Grafana
• 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
• Today, the Interactive Learning tool includes a rule-based recommendation engine that provides useful contextual recommendations
• 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
• This is an applied product data science role
• You will personally build, deploy, and operate recommendation models, design experiments, establish evaluation methodology, and define the scientific roadmap
• You will partner closely with software engineers who own the production recommender codebase and with an existing Data Analyst who supports measurement, instrumentation, and analysis across Developer Advocacy
• The long-term vision is ambitious, but we do not expect it to arrive in one release
• Evolve the Interactive Learning Plugin's recommendation system
• Develop increasingly personalized approaches to candidate selection, ranking, sequencing, and next-best-action recommendations
• You’ll own a real-time recommendation service
• Build and operate applied models
• Develop, validate, version, monitor, and iterate on models used by the recommendation system
• You’ll own model training & serving
• Define what recommendation quality means
• Develop offline, online, and longitudinal measures of recommendation performance
• You’ll own feature pipelines, monitoring of the model and architecture
• Ship incremental improvements
• Use the data and infrastructure available today while identifying the instrumentation and platform capabilities needed tomorrow
• Integrate improvements into the existing recommender rather than waiting for a complete replacement system
• Partner across disciplines
• Work closely with software engineers & data analysts to productionize models and integrate them safely into the recommender service
• Partner with the Product Analytics team on metric definitions, instrumentation, data quality, dashboards, and experiment analysis
• Collaborate with Developer Advocacy, Docs, Product, Engineering, GTM, and other teams to translate ambiguous needs into testable hypotheses and measurable product decisions
• Explain modeling choices, tradeoffs, uncertainty, and results clearly to both technical and non-technical audiences