Senior Research Scientist – Efficient Multimodal Video Generation
careers.huaweirc.ch
- Job written in
- English
- Location
- Zürich
- Work type
- On-site
- Type
- Full-time
Huawei's European Research Institute (ERI) is a global ICT leader with more than 1,500 researchers across Europe, including a Swiss laboratory that develops next‑generation AI‑centric computing architectures, hardware, software and algorithms. The institute's Multimodal Efficiency stream focuses on making video‑and‑multimodal models faster and cheaper, working hand‑in‑hand with the Architecture and Systems streams to run these models in real time on NPU/GPU clusters. The Senior Research Scientist will steer the Multimodal Efficiency research agenda, turning ideas into prototypes, large‑scale demos and deployable systems. Day‑to‑day work includes defining research directions, leading end‑to‑end projects, implementing and optimizing models, and collaborating with the Systems stream to profile and accelerate workloads on high‑performance clusters. The role also involves mentoring junior researchers and interns, publishing in top conferences, contributing to open‑source, and establishing benchmarks for quality, latency and cost. Candidates must hold a PhD in Computer Science, Electrical Engineering, AI, Machine Learning or a closely related discipline and have at least five years of post‑doctoral research experience in academia or industry. A strong record of first‑author papers at premier venues (CVPR, ICCV/ECCV, NeurIPS, ICLR, ICML, SIGGRAPH, MLSys) is required, together with hands‑on expertise in diffusion, flow‑matching or other large generative models and proven, quantifiable improvements in model efficiency through techniques such as distillation, sparse attention, token reduction or quantization. Experience leading research projects, mentoring scholars, and conducting large‑scale distributed training or inference is essential, as are solid programming skills in Python and PyTorch; C++ knowledge is a plus. Clear written and oral communication is also mandatory. Additional expertise that would strengthen an application includes work on video generation, streaming or autoregressive video models, adaptive tokenization and multi‑resolution positional encodings, step‑wise distillation, efficient attention mechanisms, Mixture‑of‑Experts models, KV‑cache optimisation, profiling on NPUs/GPUs, reinforcement‑learning or preference‑based optimisation for generative models, participation in open‑source AI projects, and experience supervising PhD students or building academic collaborations. The position offers the chance to shape a growing research stream, conduct hands‑on work at the frontier of multimodal AI and efficient video generation, and collaborate closely with leading universities, research institutions and industry partners in a vibrant, open research environment. What the role asks for: - PhD in Computer Science, EE, AI or related field - Minimum 5 years post‑PhD research experience - Strong first‑author publication record at top AI/vision venues - Hands‑on experience with diffusion or flow‑matching generative models - Demonstrated measurable efficiency gains (latency, memory, cost) - Experience leading research projects and mentoring researchers - Large‑scale distributed training or inference experience - Proficiency in Python and PyTorch - C++ programming skills (nice‑to‑have) - Experience with video generation or streaming models (nice‑to‑have) - Knowledge of token merging, adaptive tokenization or multi‑resolution encodings (nice‑to‑have) - Expertise in step distillation, sparse/efficient attention or quantization (nice‑to‑have) - Familiarity with MoE models (nice‑to‑have) - KV‑cache optimisation and long‑context modeling experience (nice‑to‑have) - Profiling/optimising models on NPUs or GPUs (nice‑to‑have) - Background in RL or preference optimisation for generative models (nice‑to‑have) - Contributions to open‑source AI projects (nice‑to‑have) - Supervision of PhD students or academic collaborations (nice‑to‑have)
Scientist: Lebenslauf-Vorlage
Bewirb dich mit einem Lebenslauf im Schweizer Aufbau — mit Beispieltext und den Anforderungen, die Scientist-Inserate am häufigsten nennen.
Lebenslauf-Vorlage Schweiz ansehenScientist: was der Markt gerade verlangt
Scientist-Stellen gehören zu den regelmässig ausgeschriebenen Berufen auf SwissJobs.app.
Weitersuchen
- Alle Wissenschaftler Jobs in der Schweiz
- Alle Jobs in Zürich
- Lebenslauf als Scientist: Beispiel und Vorlage
- Lebenslauf mit KI erstellen
- ATS-Lebenslauf prüfen
- Bewerbungsschreiben für dieses Inserat