Electra is deploying thousands of fast-charging points across Europe, and ML and GenAI are powering the products behind them. Example ML projects in production include demand forecasting for new and existing stations, GenAI assistants (ElectraDrive, our custom route planner), station recommendations, fraud detection, and dynamic pricing. Today these run as isolated prototypes with no shared infrastructure. Your mission is to get our ML and GenAI use cases into production quickly and keep them performing there. You'll collaborate with the Data Scientist early on and own the modelling that makes it production-ready: feature engineering, training pipelines, and performance optimization. Own model performance in production: monitor drift, quality, and prediction-vs-actuals; set up automated retraining, and iterate on the models when they degrade. Build the evaluation framework that gates every model rollout, especially for GenAI use cases (example: ElectraDrive). Define a standard, repeatable path from prototype to production for batch and event-driven inference: the deployment contract (standard packaging + model registry) integrating with Dagster (batch) and Temporal (event-driven, backend-owned). Set up just enough shared tooling to keep modelling consistent and repeatable: a feature store (dbt/Snowflake) for training/serving consistency and a model registry with versioning and lineage (MLFlow or similar). Contribute to the parts of our data-platform infrastructure directly tied to ML workloads: primarily orchestration (Dagster) and, where it supports ML data needs, the lake

Electra is deploying thousands of fast-charging points across Europe, and ML and GenAI are powering the products behind them. Example ML projects in production include demand forecasting for new and existing stations, GenAI assistants (ElectraDrive, our custom route planner), station recommendations, fraud detection, and dynamic pricing. Today these run as isolated prototypes with no shared infrastructure. Your mission is to get our ML and GenAI use cases into production quickly and keep them performing there. You'll collaborate with the Data Scientist early on and own the modelling that makes it production-ready: feature engineering, training pipelines, and performance optimization. Own model performance in production: monitor drift, quality, and prediction-vs-actuals; set up automated retraining, and iterate on the models when they degrade. Build the evaluation framework that gates every model rollout, especially for GenAI use cases (example: ElectraDrive). Define a standard, repeatable path from prototype to production for batch and event-driven inference: the deployment contract (standard packaging + model registry) integrating with Dagster (batch) and Temporal (event-driven, backend-owned). Set up just enough shared tooling to keep modelling consistent and repeatable: a feature store (dbt/Snowflake) for training/serving consistency and a model registry with versioning and lineage (MLFlow or similar). Contribute to the parts of our data-platform infrastructure directly tied to ML workloads: primarily orchestration (Dagster) and, where it supports ML data needs, the lake