NLP Engineer

Job Description

DataMinds Analytics is hiring an NLP Engineer for our Berlin language-technology team building multilingual text intelligence products. You will work on document understanding, sentiment analysis and conversational AI supporting 30+ languages — fine-tuning transformer models, building annotation pipelines and deploying low-latency inference services. Our Berlin team processes documents in over thirty languages for European enterprise clients, tackling low-resource languages where creative engineering makes the difference.

We sponsor EU visas, offer German courses, and give you access to serious GPU infrastructure. Our Berlin office hosts researchers and engineers from 25 nationalities, with English as the working language and a strong open-source culture. You will also collaborate with university research partners, publish at ACL workshops we sponsor, and receive funded advanced courses in multilingual LLM fine-tuning.

Responsibilities

Fine-tune transformer models (BERT, RoBERTa, LLMs) for NLP tasks
Build data annotation and active-learning pipelines
Develop multilingual text processing and evaluation systems
Deploy NLP models with optimised inference (ONNX, quantization)
Research state-of-the-art approaches and prototype rapidly
Contribute to open-source NLP tooling and publications

Requirements

3-5 years of NLP or ML engineering experience
Deep knowledge of transformer architectures and Hugging Face ecosystem
Strong Python and PyTorch skills
Experience with multilingual NLP challenges
Familiarity with model optimisation for production
Publications or open-source contributions valued

Skills

Python, PyTorch, Hugging Face, Transformers, spaCy, ONNX, Docker, Kubernetes, SQL, Evaluation Metrics

Benefits

EU visa sponsorship and relocation support
30 days holiday plus German public holidays
GPU compute access for personal research
German language courses
Conference and publication sponsorship