← All roles

Natural Language Processing Researcher

Apply on company site
Prague, Czech Republicfull-timePosted 3mo ago

About the role

On-site, full-time NLP researcher role in Prague focusing on turning AI research into real-world products, with emphasis on transformer-based models, experimentation pipelines, and rigorous evaluation.

Responsibilities

  • Take ownership of the full experimentation pipeline, including data preparation, rapidly scaling experiments from early prototypes to state-of-the-art performance, and designing rigorous, real-world relevant evaluation frameworks.
  • Explore and implement ideas to maximize performance on transformer architectures, encoders/decoders, loss functions, and optimizers.
  • Work with PyTorch or similar deep learning frameworks to build and evaluate models.
  • Communicate results effectively in research discussions and decisions.
  • Collaborate in a small, senior, high-trust team with direct access to breakthrough AI, prioritizing craftsmanship and rapid learning.

Qualifications

  • Foundational understanding of Transformer architectures, familiar with encoders, decoders, loss functions, and optimizers.
  • Hands-on experience with PyTorch or similar deep learning frameworks.
  • Ability to demonstrate novel ideas and utilize available compute for performance improvements.
  • Strong communication skills, comfortable in research discussions, and able to explain and present methods and results.
  • Experience owning the full experimentation pipeline, including data preparation and evaluation frameworks.
  • Willingness to work onsite in Prague, Czech Republic; relocation support available for EU-authorized candidates.

Skills

Required:Transformer architecturesPyTorch or similarExperimentation pipelineCommunication
Last updated Oct 20, 2026. Interested? Apply on the company site.

About BottleCap AI

Making efficiency-first LLMs & Applications

At BottleCap AI, we focus on architecture-first, efficiency-focused foundational LLMs, designed to reason better per unit of compute and unlock applications that were previously impractical. All under one roof.

efficiencyLLMsfoundational modelsapplication developmentEuropeAI research

More open roles at BottleCap AI