Data Scientist

Role summary Measure how well AI models and features perform, build predictive models where classical machine learning fits better than an LLM, and turn usage data into decisions. Responsibilities Design evaluation methods and build test datadacodataasets: expected answers, edge cases, multilingual and adversarial prompts Run evaluations and score the results with automated metrics and LLM-as-judge, calibrated against human review Benchmark models on quality, latency and cost, and recommend which to use Analyse errors, including false positives and false negatives of safety and privacy filters Build and validate predictive and classification models on tabular, time-series and signal data Add evaluation tests to CI pipelines, so a regression blocks the releacompse Analyse usage, cost and quality data, and build reports and dashboards Technical requirements Degree in data science, statistics, computer science or a related field Python with pandas, NumPy and scikit-learn Statistics: precision, recall, confidence intervals, hypothesis testing SQL Git Nice to have More experience is a plus: professional years in data science or applied machine learning, and models used in production Experience evaluating NLP or LLM systems PyTorch or TensorFlow Time-series or signal-processing models LLM evaluation tools: Langfuse, Ragas, promptfoo Experiment tracking with MLflow or Weights & Biases Experience mentoring other data scientists A second language, or a background in healthcare or another regulated domain Degree in data science, statistics, computer science or a related field Python with pandas, NumPy and scikit-learn Statistics: precision, recall, confidence intervals, hypothesis testing SQL Git