Вакансия из Hirify Global, списка международных tech-компаний Для мэтча и отклика нужен Plus
Пожаловаться
Зарплата и рынок
ЗП не указанаразвернуть
На похожих ролях по рынку
$190к$211кмедиана$269к
50 в выборке
На СНГ рынке: $5.3к/мес ($3.9к - $6к)
79
Хорошая вакансия
развернуть
Зона ответственности за production-системы рекомендаций на базе LLM описана четко, а стек современный. При этом быстрый темп клиентских проектов и формат аутсорсинга могут означать смену приоритетов.
Быстрый темп
Современный стек
Полный цикл
Сфера AI
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Подключить Plus
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Staff ML Data Scientist (LLM Recommender Systems) (PyTorch/AWS): Building, training, and deploying large-scale LLM-based recommender systems from traveler and catalog data through production inference with an accent on distributed training, fine-tuning, ranking, and privacy-safe customer data processing. Focus on designing feature and context representations, optimizing low-latency inference, evaluating model behavior, and solving data leakage and scalability challenges across tens of millions of documents.
Location: Remote in Canada or the United States
Company
Robots & Pencils designs and delivers production-ready AI systems for enterprise operations, combining engineering with creative product development.
What you will do
Transform structured and unstructured traveler and catalog data into narrative documents and synthetic data at a scale of tens of millions of documents.
Design feature and context engineering strategies, including market-relative price, tier, and trip context.
Run distributed training on Amazon SageMaker AI and HyperPod, and fine-tune models with LoRA/QLoRA, supervised training, and reinforcement learning.
Develop deep learning architectures, ranking heads, co-training, and multi-task setups for recommender systems.
Deploy low-latency, cost-efficient inference using quantization, distillation, profiling, and large-scale batch inference.
Build evaluation sets, run ablations, detect data leakage, communicate findings to Expedia and AWS science stakeholders, and handle customer data securely.
Requirements
Multi-year production machine learning experience as an applied scientist; recent transition from analytics or generalist data science is not sufficient.
Experience with Spark/PySpark and the Hadoop ecosystem, including Hive and HDFS, on corpora of tens of millions of documents.
Hands-on Amazon SageMaker AI experience across processing, training, distributed training with HyperPod, and deployment.
Experience fine-tuning LLMs with LoRA/QLoRA and applying reinforcement learning methods such as GRPO or PPO-style approaches.
Strong PyTorch deep learning fundamentals, including architectures, training dynamics, and co-training.
Experience with large-scale batch and production LLM inference, such as vLLM or Amazon Bedrock batch.
Knowledge of AWS data privacy and security, including personal-data removal, IAM, KMS, and Lake Formation.
Clear written communication with customer and science stakeholders.
Demonstrable use of AI-forward tools such as Claude and Cursor.
Nice to have
AWS Trainium or Inferentia experience with the Neuron SDK.
Recommender systems expertise, including NDCG, MRR, sequential recommenders, or generative recommenders.
Embedding quantization or semantic IDs experience.
Customer-facing or embedded consulting experience.
Culture & Benefits
Work directly with a science team, the AWS account team, and the AWS Applied Scientist leading the science.
Own the full path from raw customer data to a deployed model.
Build production-ready AI systems for real enterprise workflows and measurable client impact.
Work at a rapid delivery pace, with AI systems shipped in 30 to 45 days.
Будьте осторожны
Если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого. Это мошенники. Обязательно жмите «Пожаловаться» или пишите в поддержку. Подробнее в гайде →