Senior LLM Engineer (SWE Agent Development)

Навыки

• Дообучение моделей
• LLM
• Machine Learning
• NLP
• Transformers / HuggingFace

Задачи

• Develop an agent for solving SWE tasks
• Train and fine-tune LLMs (pre-training/fine-tuning)
• Implement agent logic running inside a sandbox
• Build validation and evaluation pipelines for models and agents
• Collect and clean datasets for training
• Select architectures and pre-trained models
• Combine models/adapters to improve performance
• High Sky — a fast-moving AI startup building competitive, production-grade solutions in text, audio, video generation, data enrichment, and predictive analytics
• The team runs the full cycle: from problem analysis to deploying real-world AI systems
• Job Description
• We’re hiring a Senior LLM Engineer to develop an agent capable of solving SWE tasks
• You’ll work on training large models, building validation pipelines, selecting architectures, and delivering fully functional AI agents within sandbox environments

Требования

• 2+ years of hands-on experience training LLMs (fine-tuning / pre-training)
• Experience with multiple LLM training methods and pre-trained models
• Strong hands-on background with Hugging Face ecosystem (Transformers, Datasets, PEFT, TRL)
• Deep understanding of NLP, transformers, LLM training cycles, RL, and classical ML
• 6+ years of commercial development experience
• Experience building products from scratch
• Degree in Computer Science, Applied Math, or related fields

Будет плюсом

• Experience with SWE-bench tasks
• Research work in Computer Science
• Awards in Kaggle/Boosters competitions, hackathon participation
• Location: Remote (no country restrictions)
• Contact: @perovvaav
• Please forward the job posting when applying

Паспорт вакансии

История публикации

Откуда что взялось

Отмечено то, что вывели мы. Без пометки — значение назвал работодатель.

Почему на этом месте в выдаче

Порядок выдачи объявлен контрактом: свежесть решает между днями, полнота и зарплата — внутри дня.