Geared toward candidates comfortable with end-to-end ML model ownership and building evaluation tooling, with a focus on data-driven solutions and collaboration across technical and non-technical teams in the CEE region and EU.
Derived from job-description analysis by Serendipath's career intelligence engine.
Original posting from Poland and Middle-Eastern Europe via Greenhouse
Xebia is a global AI-first, digital transformation, and engineering partner. With over 25 years of experience and a team of 5,000 professionals across 16 countries, we help organizations design and build scalable products, platforms, and data-driven solutions.
We specialize in Artificial Intelligence, Data and Cloud, Intelligent Automation, and Digital Products, combining deep technical expertise with a strong focus on engineering excellence and a people-first culture.
In the CEE region, we’re a team of nearly 1,000 experts delivering modern applications, data platforms, and AI solutions for clients such as Millennium, ING, Play, Edenred, Arabian Drilling, FedEx, Leroy Merlin, Truecaller, Volotea, Schmitz Cargobull, and many, many more. We work with leading technologies including AWS, Azure, GCP, Databricks, and Snowflake, and combine strong engineering culture with a consulting mindset and a continuous focus on growth and knowledge sharing.
You will be:
Owning models end to end across caller identity, business identification, spam and fraud detection: problem framing, data, training, evaluation, deployment and monitoring.
Working on ranking, classification, graph, personalisation and LLM-based models.
Setting the team’s evaluation standards through historical backtests, honest baselines, false-positive checks on legitimate callers, shadow runs and A/B tests.
Building shared tooling that makes good practice the easy path, including reusable evaluation harnesses, experiment tracking, project templates and workflows for moving work from notebooks into production code.
Working with product stakeholders to turn loosely defined problems into clear modelling goals and success metrics, and explaining results and trade-offs to non-technical audiences.
Sharing responsibility for model monitoring and incident handling, including drift detection.
Documenting decisions and results in short, reproducible analyses.
Your profile:
Several years of hands-on experience building and shipping machine-learning models to production.
Strong Python and SQL skills, with experience working with large-scale data in BigQuery or a similar data warehouse.
Experience with classical machine learning, such as gradient boosting and ranking, and deep learning using TensorFlow/Keras or PyTorch, including training on GPUs.
Experience designing offline evaluations that predicted online results, and running A/B tests.
Experience with data and ML pipelines using Airflow, dbt and Spark, or with running training jobs on Kubernetes.
Work from the European Union region and a work permit are required.
Nice to have:
Experience gathering requirements from non-technical stakeholders, turning them into a scoped plan, and communicating results and trade-offs clearly.
Experience setting or improving data science standards within a team, including evaluation, reviews and reproducibility.
Experience with graph machine learning, such as GNNs, link prediction or label propagation, or with recommendation and retrieval models, such as two-tower models and embeddings.
Experience using LLMs in production for extraction, classification or generation, including LLM-as-judge evaluation.
Familiarity with MLOps tools such as MLflow for experiment tracking, model registries, versioning and reproducible training.
Experience with NLP on short, multilingual or transliterated text, such as names, or computer vision on user images.
Experience with telecom or calling data, or in fraud detection.
Recruitment Process:
CV review – HR call – Interview – Client Interview – Decision
Source: Poland and Middle-Eastern Europe careers (Greenhouse)