AI Researcher

Role overview

This opportunity is being handled through a partner company, which will manage all applications and the following stages of the hiring process. The employer is seeking an AI Researcher located in Germany.

The position is centered on advanced agentic AI research, with a focus on creating next-generation systems that improve from large volumes of real-world interaction data. The work involves turning multimodal inputs such as text, audio, logs, and structured workflows into stronger reasoning and decision-making models. It combines deep research with practical delivery, so discoveries are expected to move into live systems and influence real user-facing agents. You will work closely with engineering and product counterparts to help research results reach production and improve continuously from live feedback. This is a role for someone who enjoys working through open-ended problems and wants to contribute to the evolution of multimodal and agentic AI.

Key accountabilities

• Carry out advanced research on agentic AI systems trained on real-world interaction data, with emphasis on better reasoning, planning, and tool usage.
• Build and test learning approaches such as RAG, fine-tuning, RLHF, DPO, and GRPO to raise the performance of large-scale models.
• Create multimodal representation learning methods, including shared embedding spaces for text, audio, logs, and structured data.
• Advance speech and audio intelligence capabilities, covering STT, ASR, and audio-based learning workflows.
• Establish evaluation methods to assess agent performance in real-world and domain-specific settings.
• Convert behavioral and interaction signals into structured training goals for large models.
• Partner with engineering and product teams to operationalize research and refine it using live system feedback.

Requirements

• A PhD in Computer Science, Machine Learning, AI, Electrical Engineering, or a closely related discipline.
• At least 5 years of experience in applied AI research or ML systems that have had production impact.
• Deep knowledge of large-scale machine learning, LLMs, or multimodal AI systems.
• Practical experience with RAG, LLM fine-tuning, and reinforcement learning approaches such as RLHF, DPO, or GRPO.
• Solid foundation in representation learning, embeddings, and multimodal joint spaces.
• Exposure to speech and audio modeling, including STT, ASR, or audio signal processing.
• Strong Python skills and familiarity with modern ML tools such as PyTorch and Hugging Face.
• Experience building evaluation frameworks for LLMs or agentic systems.
• Ability to form research hypotheses from ambiguous real-world challenges.
• Excellent written and spoken English communication skills.

Benefits

• Completely remote role within a global AI research organization.
• Chance to help shape advanced agentic and multimodal AI systems.
• Research work with direct impact on production deployment.
• Exposure to high-caliber engineering and product collaboration.
• Strong ownership of research direction and technical strategy.
• Access to large proprietary datasets and AI infrastructure.
• Competitive pay based on experience and expertise.

Application and privacy information

The application process is managed by an AI-assisted matching system designed to review candidates quickly, consistently, and fairly against the core needs of the role. A shortlist of top-matching applicants is shared with the hiring company, while interviews, assessments, and final next steps are handled internally by the employer.

By applying, you consent to the processing of your personal data for candidacy evaluation and for sharing relevant details with the hiring employer. This is done under legitimate interest and pre-contractual measures where applicable, including GDPR-related rights such as access, correction, deletion, and objection.

AI tools may be used to support parts of the recruitment workflow, including resume review, application analysis, response evaluation, and verification of possible inconsistencies. These tools assist the recruitment team but do not replace human judgment, and the final hiring decision is made by people. Further information about data processing can be requested directly.