Geforderte Skills
- Graph Neural Networks
- GraphStorm
- Amazon SageMaker
- AWS CloudFormation
- Pulumi
- MLOps
- Data Engineering
- Machine Learning Pipelines
- SageMaker
- Machine Learning
- Data Science
- Cloudformation
Sprachen: English
Jobbeschreibung
Project overview
Data Scientist / ML Engineer (AWS, Graph Neural Networks) – 90% remote, 3-year engagement (01.01.2027 – 31.12.2029, approx. 220 days/year).
For a large German mobility corporation (IT subsidiary) we are looking for up to 3 freelance Data Scientists / ML Engineers based in the EU or Germany. The project develops, productionizes and continuously optimizes scalable ML models for fraud detection in online ticket sales and for optimizing operational review processes, within an agile team on AWS.
Responsibilities
- Build and operate scalable data pipelines (Python, PySpark, Delta Lake, Parquet, polars/pandas; AWS S3 and Azure Blob Storage)
- Data preparation, feature engineering and graph construction from tabular data for Graph Neural Networks (GraphStorm, DGL, PyTorch)
- Graph databases on Amazon Neptune / Gremlin (data modelling, bulk loading); relational databases on Amazon Aurora / RDS / PostgreSQL
- Automated ML pipelines on AWS SageMaker (processing & training jobs, hyperparameter tuning, Model Registry, endpoints)
- Productive ML components with GNNs and XGBoost, incl. explainability (SHAP)
- Data and inference interfaces with AWS Lambda, API Gateway, Step Functions, event-driven processing with Kafka/MSK
- Infrastructure as Code (CloudFormation, Pulumi) and GitLab CI/CD
- Integration of AI agent systems (MCP, Langfuse), dashboards with Grafana and Plotly
Requirements
Nice-to-have: Kafka/AWS MSK, XGBoost, SHAP, AI agents / MCP / Langfuse, PySpark / Delta Lake, S3 & Azure Blob integration.
Language: English C1 (no German required).
- Production experience with GraphStorm or DGL incl. graph construction
- 3+ years of AWS SageMaker (pipelines, processing/training jobs, Model Registry, endpoints)
- Professional experience in Data Engineering and/or MLOps with production data/ML solutions
- Infrastructure as Code with AWS CloudFormation or Pulumi
- Automated pipelines for structured and unstructured data
Project details
remote from the EU/Germany, occasional onsite workshops in Frankfurt am Main (approx. 20 days per year).
- Start: 2027-01
- Remote: hybrid
- Location: Frankfurt am Main, Deutschland, DE
- Languages: English
Indexiert von der 9am-Meta-Suche. Du bist Job-Anbieter? Listing beanspruchen oder verwalten
Gefunden auf FreelancerMap DE