Research Engineer: Equity & Production-Ready AI Pipelines

Wwshemi; Zürich; Hybrid; CHF 120.000 - 180.000; Vollzeit

Benefits dieser Stelle

Equity provided

Visa sponsorship

Zusammenfassung

Wwshemi is seeking a Research Engineer to turn experimental code into production-grade systems within an AI research team. You will build shared infrastructure and post-training pipelines to scale research projects while preserving experimental integrity.

You will implement end-to-end data generation, training and evaluation pipelines, and distributed experiment infrastructure. You’ll profile workloads, improve GPU utilization and data throughput, and collaborate with researchers to translate

Qualifikationen

• Turn research prototypes into tested, reusable data generation, training, and evaluation pipelines.
• Experience building and debugging software for model training, inference, or large-scale data processing.
• Understanding of ML experiments and how data, precision, and implementation choices affect results.

Aufgaben

• Turn research prototypes into tested, reusable data generation, training, and evaluation pipelines.
• Build distributed experiment infrastructure for data loading, checkpointing, and collecting agent interactions.
• Profile workloads and improve GPU utilization, memory efficiency, and data throughput.
• Develop tests, experiment tracking, and debugging tools while ensuring changes preserve research results.
• Work with researchers to translate ideas into robust, maintainable software.

Kenntnisse

Python

ML Frameworks (PyTorch/JAX)

Model training software

Experiment design

Testing & profiling

Distributed training tools (PyTorch+De

Jobbeschreibung

Wwshemi is seeking a Research Engineer to turn experimental code into production-grade systems within an AI research team. You will build shared infrastructure and post-training pipelines to scale research projects while preserving experimental integrity.

You will implement end-to-end data generation, training and evaluation pipelines, and distributed experiment infrastructure. You’ll profile workloads, improve GPU utilization and data throughput, and collaborate with researchers to translate