(Senior) AI Researcher (LLM)

Location: Hybrid in Heidelberg or Berlin, Germany

Company

hirify.global develops sovereign, human-centric AI and foundational models for governmental institutions and large enterprises, with applications in development, engineering, logistics, and manufacturing.

What you will do

• Design, execute, and scale supervised fine-tuning and reinforcement learning methods to reduce LLM hallucinations and improve evidence grounding.
• Research uncertainty estimation, calibration, self-consistency, and self-correction mechanisms that help models identify when they do not know.
• Build and optimize high-throughput distributed fine-tuning and training pipelines across multi-node GPU clusters.
• Design rigorous evaluations for hallucination rates, truthfulness, and reasoning accuracy.
• Drive research projects independently from initial hypotheses through scalable production implementation and technical reporting.

Requirements

• Hands-on experience fine-tuning or post-training large language models with modern distributed training frameworks.
• Research or implementation experience in hallucination mitigation, truthfulness alignment, grounding in retrieval-augmented generation, uncertainty estimation, out-of-distribution detection, RAG evaluation, refusal calibration, or claim verification.
• Strong programming skills in Python and PyTorch.
• Availability for hybrid work in Heidelberg or Berlin, Germany.

Culture & Benefits

• Work on foundational model research and next-generation AI methods with access to hundreds of GPUs.
• 30 days of paid vacation.
• Flexible working hours and a hybrid working model.
• Fitness and wellness offerings, including Wellhub, plus mental health support through nilo.health.
• Subsidized company pension plan and Germany-wide transportation ticket.
• Budget for additional technical equipment and a subsidized JobRad bike lease.