Research (AI)

TL;DR

Research (AI): Stress-testing and evaluating frontier LLMs and autonomous agents with an accent on benchmarking and identifying model-breaking points. Focus on creating benchmark data, designing experimental research, and interpreting linguistic limits of modern AI.

Location: Hybrid (Berlin, Germany). Must be currently enrolled at TU Berlin

Company

AI-driven translation platform specializing in machine translation and human-in-the-loop expertise for global enterprises.

What you will do

• Run rigorous evaluations on frontier LLMs and autonomous agents across diverse tasks.
• Create and modify benchmark data to test the reasoning and linguistic limits of modern AI.
• Design and run experiments to identify model-breaking points and interpret the resulting data.
• Work under the supervision of in-house research staff to improve model capabilities.

Requirements

• Currently enrolled at TU Berlin majoring in Computer Science (Bachelor/Master) or a related field.
• Solid understanding of LLMs, natural language processing, or machine learning.
• Highly proficient in Python, Bash, and git.
• Proficient in English.
• Strong drive to ship high-quality projects, sometimes on tight deadlines.

Nice to have

• Proficiency in one or more non-English languages.

Culture & Benefits

• Direct collaboration with models and teams from frontier AI labs.
• Opportunity to publish papers in top-tier AI/ML conferences.
• Contribution to industry-standard open-source benchmarks.
• Competitive salary.
• Hybrid work environment with an on-site research team in Berlin.