Berlin ML Research Engineer Intern

United States Digital Space LLC; United States; Remote; USD 12,000 - 19,000; Full time

Job summary

United States Digital Space LLC is seeking a driven intern for our Berlin office to contribute to search and retrieval research. You will train and optimize large-scale deep learning models using PyTorch with distributed training and hardware acceleration, focusing on retrieval, multilingual and multimodal modeling.

You will build and evaluate RAG pipelines for grounding and answer generation, with opportunities to present research at AI/ML venues.

Qualifications

• Understanding of search and retrieval systems with evaluation metrics.
• Strong PyTorch proficiency including distributed training and optimization.
• Interest in representation learning, contrastive methods and multilingual vectors.
• Publication record in AI/ML conferences/workshops (e.g., NeurIPS, ICML, ICLR).

Responsibilities

• Relentlessly push search quality forward through models, data, tools, or other leverage.
• Train and optimize large-scale deep learning models using PyTorch with distributed training and acceleration.
• Conduct research in representation learning, including contrastive learning and multilingual, evaluation, and multimodal modeling for search.
• Build and optimize RAG pipelines for grounding and answer generation.

Skills

Search and retrieval systems

PyTorch distributed training

Representation learning

Academic publications

Tools

PyTorch

DeepSpeed

FSDP

Job description

United States Digital Space LLC is seeking a driven intern for our Berlin office to contribute to search and retrieval research. You will train and optimize large-scale deep learning models using PyTorch with distributed training and hardware acceleration, focusing on retrieval, multilingual and multimodal modeling.

You will build and evaluate RAG pipelines for grounding and answer generation, with opportunities to present research at AI/ML venues.