AI Researcher — Distillation

Jobgether; Deutschland; Vor Ort; EUR 90.000 - 150.000; Vollzeit

Benefits dieser Stelle

Ownership over research direction

Support for publishing research

Access to compute resources

Collaborative, technical team

Opportunities to deploy research

Zusammenfassung

Jobgether, on behalf of a partner company, is seeking an AI Researcher focused on distillation in Germany. Join a highly technical environment to transform large models into smaller, faster, deployable systems while maintaining quality.

You will explore teacher-student distillation, long-context architectures, and inference-constrained deployment, moving ideas into production. You will publish work, contribute to open-source projects, and collaborate with engineers to productionize approaches.

Qualifikationen

• Strong academic or professional background in machine learning research.
• Hands-on experience with model distillation or related areas.
• Demonstrated publication experience through conference or journal papers.

Aufgaben

• Design, implement, and evaluate advanced model distillation techniques.
• Investigate tradeoffs between model size, latency, memory, throughput, and accuracy.
• Develop approaches for distilling large language models and long-context architectures.
• Conduct large-scale experiments and rigorous analysis to validate hypotheses.
• Translate research findings into production-ready implementations with engineering teams.
• Prepare and submit research papers to leading ML conferences.

Jobbeschreibung

This position is listed on behalf of a partner company, who manages all applications and next steps. Our partner is looking for a AI Researcher — Distillation based in Germany.

Join a highly technical research environment focused on advancing the efficiency and performance of modern AI models.

You will research techniques that transform large, resource-intensive models into smaller, faster, and more deployable systems without sacrificing quality.

The role combines fundamental machine learning research with hands‑on experimentation and real‑world engineering challenges.

You will have the opportunity to explore model distillation across large language models, long‑context architectures, and inference‑constrained environments.

Your work will move from research ideas and rigorous experiments into production systems with measurable impact.

You will collaborate closely with engineers while contributing to publications, technical research, and potentially open‑source projects.

This opportunity is particularly suited to researchers who want meaningful ownership of their work and the ability to see their ideas deployed in practice.

Accountabilities

• Design, implement, and evaluate advanced model distillation techniques, including teacher-student training, self‑distillation, layer‑wise distillation, and representation matching.
• Investigate the tradeoffs between model size, latency, memory consumption, throughput, and accuracy.
• Develop novel approaches to distilling large language models, long‑context or specialized architectures, and models designed for inference‑constrained environments.
• Conduct large‑scale experiments, ablation studies, and rigorous analysis to validate research hypotheses and identify meaningful improvements.
• Translate research findings into practical implementations and collaborate closely with engineering teams to productionize successful approaches.
• Prepare and submit research papers to leading machine learning conferences and venues such as NeurIPS, ICML, ICLR, and COLM.
• Contribute to internal research documentation, technical articles, and open‑source machine learning projects where appropriate.
• Clearly communicate research objectives, methodologies, results, tradeoffs, and limitations to technical stakeholders.

Requirements

• Strong academic or professional background in machine learning research, with a solid understanding of deep learning fundamentals.
• Hands‑on experience with model distillation or closely related areas such as model compression, pruning, quantization, or representation learning.
• Demonstrated publication experience through conference or journal papers, workshop publications, or arXiv preprints.
• Strong understanding of optimization, training dynamics, generalization, and modern deep learning methodologies.
• Fluency in PyTorch or an equivalent deep learning framework, with experience conducting research‑grade experimentation.
• Ability to design rigorous experiments, interpret results, and critically evaluate research approaches.
• Strong written and verbal communication skills, with the ability to explain complex research ideas and findings clearly.
• Experience with large language model distillation is highly valued.
• Background in efficiency‑focused research involving latency, memory, throughput, or related deployment constraints is advantageous.
• Experience with long‑context models or non‑Transformer architectures is a plus.
• Open‑source contributions to machine learning, research tooling, or related projects are beneficial.
• Prior startup or applied research experience is welcome.
• PhD, postdoctoral, academic research, or industry research experience in machine learning or a related field is particularly relevant, though equivalent research backgrounds may also be considered.

Benefits

• Significant ownership and influence over research direction within a Series A‑stage environment.
• Strong support for publishing research and pursuing open research initiatives.
• Close feedback loop between research experimentation and real‑world production deployment.
• Access to meaningful compute resources and production‑scale machine learning problems.
• Opportunity to work on cutting‑edge model efficiency and distillation challenges.
• Collaboration within a small, highly technical team with deep expertise across machine learning and systems.
• Opportunity to see research progress from papers and experimental code through to deployed AI systems.
• Exposure to large language models, efficient inference, long‑context architectures, and other emerging AI technologies.

How Jobgether works

We use an AI‑powered matching process to ensure your application is reviewed quickly, objectively, and fairly against the role's core requirements. Our system identifies the top‑fitting candidates, and this shortlist is then shared directly with the hiring company. The final decision and next steps (interviews, assessments) are managed by their internal team.

We appreciate your interest and wish you the best!

Data Privacy Notice: By submitting your application, you acknowledge that Jobgether will process your personal data to evaluate your candidacy and share relevant information with the hiring employer. This processing is based on legitimate interest and pre‑contractual measures under applicable data protection laws (including GDPR). You may exercise your rights (access, rectification, erasure, objection) at any time.