Senior Post-Training AI Engineer for Scalable Systems

Embodied AI; Lausanne; Hybrid; CHF 150.000 - 210.000; Vollzeit

Zusammenfassung

Giotto.ai is a Switzerland-based AI company building intelligence systems for Switzerland and Europe. We are seeking a Senior Research Engineer or Research Scientist to own the training and optimisation side of our complete post-training stack.

Starting from pretrained checkpoints, you will design, implement, scale, and operate methods to produce capable, reliable production models. Your scope includes supervised fine-tuning, preference optimisation, RLHF, reward models, and multi-stage

Qualifikationen

• Experience with large-scale language-model training or post-training runs across multiple machines and accelerators.
• Experience with workloads requiring more than single-node training or data-parallelism.
• Strong Python and PyTorch expertise, with distributed execution and mixed precision.

Aufgaben

• Own the end-to-end post-training pipeline from pretrained checkpoint to production candidate.
• Design and execute full-parameter and parameter-efficient SFT.
• Implement preference optimisation, RLHF, RLAIF, and related methods.
• Develop training strategies for reasoning, coding, tool use, multilingual behaviour, and long-horizon tasks.
• Integrate reward models, verifiers, and graders into training loops.
• Scale training across multiple machines and accelerators.
• Profile and improve accelerator utilisation and end-to-end training time.
• Collaborate with data, evaluation, infra, and inference teams.

Kenntnisse

Python

PyTorch

Experiment design

Distributed training

Tools

PyTorch Distributed

FSDP

DeepSpeed

Megatron-Core

Hugging Face Transformers

CUDA

Jobbeschreibung

Giotto.ai is a Switzerland-based AI company building intelligence systems for Switzerland and Europe. We are seeking a Senior Research Engineer or Research Scientist to own the training and optimisation side of our complete post-training stack.

Starting from pretrained checkpoints, you will design, implement, scale, and operate methods to produce capable, reliable production models. Your scope includes supervised fine-tuning, preference optimisation, RLHF, reward models, and multi-stage