Senior / Data Scientist (Model Training and Evaluation) , Giant Robots Labs
Core
Building an in-house capability to evaluate, select, adapt, and post-train AI models for government use, establishing repeatable evaluation standards and training recipes.
Role type
Senior Data Scientist (Model Training and Evaluation)
Builds
Evaluation suites, model hosting infrastructure, and post-trained open-weight models for government tasks.
Domain
Public Sector / Government Technology / AI & Machine Learning
Deliverable
production ML models
Required skills
Python, PyTorch, Hugging Face Transformers, supervised fine-tuning, LoRA/QLoRA, DPO, evaluation methodology, experiment design, dataset curation, GPU infrastructure, inference serving, vLLM, SGLang
Preferred skills
Multi-GPU training (FSDP/DeepSpeed), RLHF/PPO/GRPO, model distillation, LLM-as-judge pipelines, synthetic data generation, data governance
Responsibilities
Design and build evaluation suites to compare models on government-relevant tasks; Diagnose model weaknesses through observational and controlled evaluations; Set up and operate model hosting and evaluation infrastructure; Prepare authorised training data including curation and quality checks; Run post-training experiments such as supervised fine-tuning and preference optimisation; Evaluate trained models against untrained baselines and document training recipes; Work with government agencies to assess data readiness and assemble training examples.
