Senior Machine Learning Engineer

Zendesk Zendesk · Enterprise · Poland · Remote

Senior ML Engineer to join the AI Copilot organization, owning the delivery of ML-powered product features at scale, from prototype to production. Responsibilities include building and maintaining ML infrastructure, integrating LLMs, and improving system reliability and performance.

What you'd actually do

  1. Own and deliver ML-powered product features end-to-end — from data pipelines and model integration through serving, monitoring, and iteration in production.
  2. Work closely with Scientists to productionise research outputs into reliable, user-facing features.
  3. Build and maintain ML infrastructure: model serving, inference pipelines, LLM integrations, and evaluation frameworks.
  4. Contribute to technical design discussions and architecture decisions within your team, with growing influence across teams.
  5. Collaborate with product software engineers to ensure ML capabilities are well-integrated into the broader product experience.

Skills

Required

  • Python
  • ML engineering
  • MLOps
  • building ML-powered products
  • model serving
  • inference pipelines
  • monitoring
  • integrating LLMs
  • prompt engineering
  • evaluation
  • multi-provider setups
  • SQL
  • data infrastructure
  • data pipelines
  • data quality
  • containerized deployments
  • Docker
  • Kubernetes
  • cloud infrastructure
  • AWS
  • owning features end-to-end
  • delivering to production
  • collaboration skills

Nice to have

  • Ruby
  • Snowflake
  • dbt
  • Metaflow
  • MLflow
  • BentoML
  • PyTorch
  • TensorFlow
  • Kafka
  • A/B testing
  • feature flags
  • incremental rollouts

What the JD emphasized

  • 5+ years of experience in software engineering, with a meaningful focus on ML engineering, MLOps, or building ML-powered products.
  • Solid experience building and operating ML systems in production: model serving, inference pipelines, and monitoring.
  • Experience integrating LLMs into production systems — prompt engineering, evaluation, or multi-provider setups.
  • A track record of owning features end-to-end and delivering them to production with high quality.

Other signals

  • AI Copilot is a multi-million ARR product
  • own the delivery of ML-powered product features at Zendesk scale
  • taking capabilities from prototype through to production
  • deliver early, deliver often, and iterate based on real-world customer feedback