Senior AI Engineer – Pre-training Data (f/m/d)
Core
Designing, building, and maintaining infrastructure to source, process, filter, and curate high-quality pre-training corpora for foundation models, with a focus on German-language data.
Role type
Senior AI Engineer (Pre-training Data)
Builds
Data pipelines, data quality tooling, and data mixtures for model pre-training
Domain
AI/ML, Foundation Models, Data Engineering
Deliverable
production ML models
Required skills
Python, data engineering, deep learning frameworks, workflow orchestration, object storage, columnar data formats, distributed processing, data quality analysis, ablation study design, data lineage tracking
Preferred skills
Large-scale data processing, corpus sourcing/curation, data quality methods (classifier-based filtering, heuristic scoring), foundation model training theory, web-scale data sourcing/crawling, Rust, Kubernetes, cloud-native ML infrastructure
Technologies
Python, Rust, Kubernetes, Common Crawl, WARC, object storage, columnar data formats
Responsibilities
Design and maintain end-to-end data pipelines for sourcing and preparing pre-training corpora; Curate and compose data mixtures balancing domains, languages, and quality; Build classifiers and frameworks to measure and enforce data quality; Identify data gaps and source/generate data to address model weaknesses; Collaborate with post-training teams to align data choices with downstream goals; Ensure deep coverage of German-language corpora; Design and run ablation studies to validate data choices; Maintain data lineage and provenance
Seniority
Senior, hands-on IC