Remote Senior ML Engineer — AI Reasoning & Training

Alignerr Corp.; España; Remote; EUR 74,000 - 135,000; Part time

Job summary

Alignerr is seeking a Senior Machine Learning Engineer (AI Training) for a fully remote hourly contract. You will craft high‑fidelity reasoning traces that teach how AI should plan, reason, and decide across complex real‑world tasks.

The role is fully remote with 10–40 hours per week and collaboration with a global ML team. The candidate is an experienced ML engineer, capable of breaking down ambiguous problems into structured reasoning, with deep knowledge of LLMs, their training dynamics, and

Qualifications

• Experienced ML engineer with focus on model reasoning, training, or evaluation.
• Ability to break down complex problems into structured steps.
• Deep familiarity with how LLMs work, including failure modes and training dynamics.
• Clear, precise written communicator who can articulate complex reasoning.
• Self-directed and productive in asynchronous environments.

Responsibilities

• Author high-fidelity reasoning traces showing how AI should plan, use tools, and decide across complex tasks.
• Decompose complex problems into clear, logical, well-documented steps for model training ground truth.
• Review and mentor quality of structured traces for accuracy and consistency.
• Design data strategies to help LLMs navigate multi-step decisions.
• Collaborate with a global team of ML researchers on frontier AI projects.

Skills

ML Engineering

LLMs

AI Training

Model Reasoning

Data Annotation

Job description

Alignerr is seeking a Senior Machine Learning Engineer (AI Training) for a fully remote hourly contract. You will craft high‑fidelity reasoning traces that teach how AI should plan, reason, and decide across complex real‑world tasks.

The role is fully remote with 10–40 hours per week and collaboration with a global ML team. The candidate is an experienced ML engineer, capable of breaking down ambiguous problems into structured reasoning, with deep knowledge of LLMs, their training dynamics, and