pythonllmragpytorchfine tuningprompt engineeringreinforcement learninghuggingfacepeftagentic ai
Вакансия из Hirify Global, списка международных tech-компаний Для мэтча и отклика нужен Plus
Пожаловаться
Зарплата и рынок
развернуть
По рынку СНГ данных пока нет
На международном рынке: $174к/год ($125к - $278к)
81
Хорошая вакансия
развернуть
Это хорошо описанная LLM-роль с фокусом на внедрение и актуальные AI-задачи, но широкий охват от исследований до продакшена и удалённая работа только из Украины ограничивают доступность вакансии.
?Широкий охват
Современный стек
Фокус на AI
?Только Украина
Понятные льготы
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Подключить Plus
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Applied Scientist (LLM) (Generative AI): Developing production-ready LLM features for Cloud and Edge environments with an accent on model compression, quantization, RAG architectures, and multi-agent systems. Focus on fine-tuning, synthetic data generation, safety and alignment validation, inference optimization, and benchmarking complex LLM workflows.
Location: Kyiv, Lviv, or remote within Ukraine only.
Company
hirify.global is building Generative AI features for Cloud and Edge environments through a distributed research team.
What you will do
Develop high-performance Generative AI features for private, real-time Edge inference and scalable Cloud deployment.
Curate large-scale text and multimodal training data, generate synthetic datasets, and build robust data pipelines.
Optimize inference speed and accuracy using model compression, quantization, PEFT, LoRA, QLoRA, and prompt engineering.
Evaluate hallucinations, safety, alignment, and model quality using industry benchmarks and internal KPIs, producing technical reports.
Collaborate with Product and Data Engineering teams and mentor junior colleagues.
Requirements
3+ years of commercial Machine Learning experience focused on NLP or LLMs.
Strong knowledge of Python3, NumPy, pandas, modern text-processing libraries, PyTorch, and Hugging Face Transformers, PEFT, and Accelerate.
Proficiency in PEFT/LoRA and reinforcement learning techniques.
Deep understanding of attention mechanisms, tokenization, context window management, and embedding spaces.
Practical experience with at least one of RAG, fine-tuning, or agentic frameworks.
Experience managing and analyzing datasets larger than 100GB across text, image, and audio formats, including high-fidelity dataset creation and data pipeline development.
Expertise in prompt engineering, agentic framework design, and LLM pipeline orchestration.
Experience deploying LLMs in production with Triton Inference Server, vLLM, TGI, or ONNX.
Good written and spoken English.
Nice to have
Experience with Pinecone, Weaviate, Milvus, or Chroma; LangChain, LlamaIndex, or AutoGen; and RAGAS, DeepEval, or G-Eval.
Knowledge of advanced quantization methods such as GGUF, AWQ, or EXL2, as well as pruning and knowledge distillation.
Basic understanding of web/client-server architecture and streaming API responses using Asyncio or aiohttp.
Experience with Docker, Kubernetes, and cloud GPU orchestration such as Run:ai or Lambda Labs.
Knowledge of C++, Triton, or CUDA for custom kernel development.
Culture & Benefits
Gig-contract with competitive compensation and perks.
Remote working mode is available within Ukraine only; office-based work includes free meals, fruits, and snacks.
21 paid vacation days per year and paid public holidays according to Ukrainian legislation.
Medical insurance from day one, with sick leaves and medical leaves available.
Corporate courses, knowledge hubs, free English classes, educational leaves, an equal-opportunity environment, and a transparent, value-based culture.
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 →