About the company
At Preply, we’re all about creating life-changing learning experiences. We help people discover the magic of the perfect tutor, craft a personalised learning journey, and stay motivated to keep growing. Our approach is human-led, tech-enabled - and it’s creating real impact.
We’ve just reached unicorn status with a $150M Series D, accelerating our vision to transform education through human-led, AI-enhanced learning. Today, 100,000+ tutors teach 90+ languages to learners in 180 countries - and we’re only getting started.
Responsibilities
- Build and maintain ML pipelines for training, evaluation, and deployment using tools like Databricks, MLFlow, Airflow, DBT, Sagemaker, Tecton.
- Support AI scientist creating reproducible, containerized model training environments (on-demand and scheduled), and manage compute at scale (e.g., spot/GPU autoscaling).
- Define and implement observability and alerting for ML systems (model drift, data quality, feature coverage, etc.).
- Design and scale data ingestion and feature transformation flows using batch (e.g., Spark/BigQuery) and streaming (Kafka or equivalent).
- Contribute to internal Python libraries and platform tooling that accelerate experimentation and deployment for all model teams.
- Ensure ML services are modular, testable, and monitored from day one.
- Exploration and productionization of LLM-based features (e.g., retrieval pipelines, prompt evaluation, model serving).
Requirements
- Proven experience designing and deploying ML systems in production (5+ years in relevant roles).
- Proficiency in Python and SQL, and orchestration tools (Airflow, Kubeflow, Dagster, etc.).
- Experience with modern cloud platforms (preferably GCP or AWS), Kubernetes, and CI/CD workflows.
- Understanding of ML model lifecycles: training, validation, deployment, and monitoring.
- Strong DevOps practices: Git, IaC (Terraform), logging/observability, containerization (Docker/K8s).
- Ability to work independently with ML Scientists and mentor peers in reliability, testing, and delivery.
- Exposure to LLM serving, vector databases, or GenAI-powered product flows.
- Deep, hands-on expertise in AI tools, especially in agentic AI SDLC.