Data Science Engineer
Core
Build high-quality datasets and experiments to identify data contributions to language model training, improving evaluation quality and model iteration.
Role type
Senior Data Science Engineer (LLM Data & Experiments)
Builds
Large-scale data pipelines, synthetic data generation, and language model training datasets for Mercado Libre's services.
Domain
E-commerce / Generative AI / Large Language Models
Deliverable
production ML models
Required skills
Python, SQL, text data engineering, NLP, LLM architectures, fine-tuning, evaluation, deep learning, hypothesis formulation, metric definition, technical mentorship
Preferred skills
Experience with open-source and commercial models, Hugging Face, vLLM, autonomous work, guiding teams
Technologies
Python, SQL, Hugging Face, vLLM
Responsibilities
Define data needs for models and lead end-to-end data initiatives; Design and implement scalable pipelines for data curation and synthetic data generation; Design and prioritize ablation experiments to measure data impact; Analyze results to propose iterations and guide the team via mentorship and code reviews.
