Senior ML Engineer: LLM Systems & Evaluation

Grid Dynamics; España; On-site; PHP 1,800,000 - 2,600,000; Full time

Job summary

Grid Dynamics seeks a Senior Machine Learning Engineer focused on LLM systems and evaluation. You will own ML projects from problem definition through implementation and collaborate with product, engineering, and research teams.

You will design evaluation methodologies, build datasets/benchmarks, and iterate to translate business goals into measurable ML objectives. Strong Python and ML framework experience (PyTorch, TensorFlow, or JAX) are essential.

Qualifications

• 5+ years in Machine Learning Engineering or related field.
• Strong understanding of ML fundamentals and model evaluation.
• Experience with PyTorch, TensorFlow, or JAX.
• Experience with LLMs, RAG, and safety concepts is desirable.
• Ability to work in ambiguous problem spaces with limited data.

Responsibilities

• Own machine learning projects from problem definition through implementation.
• Design and implement evaluation methodologies for AI systems.
• Create datasets, benchmarks, and metrics to measure model performance.
• Evaluate and improve LLM-based systems including RAG apps and agents.
• Build and maintain ML pipelines, tooling, and infrastructure.
• Collaborate with product, engineering, and research teams to translate goals into ML objectives.
• Prototype rapidly to solve business and product challenges.

Skills

Python programming

ML fundamentals

Model evaluation

Strong communication

Ambiguous problem solving

Tools

PyTorch

TensorFlow

JAX

Job description

Grid Dynamics seeks a Senior Machine Learning Engineer focused on LLM systems and evaluation. You will own ML projects from problem definition through implementation and collaborate with product, engineering, and research teams.

You will design evaluation methodologies, build datasets/benchmarks, and iterate to translate business goals into measurable ML objectives. Strong Python and ML framework experience (PyTorch, TensorFlow, or JAX) are essential.