Назад
Company hidden
3 дня назад

Data Scientist (NLP/Search)

Формат работы
hybrid
Тип работы
fulltime
Английский
c1
Страна
UAE/Netherlands/Turkey +1 еще
Вакансия из списка Hirify.GlobalВакансия из Hirify Global, списка международных tech-компаний
Для мэтча и отклика нужен Plus

Мэтч & Сопровод

Для мэтча с этой вакансией нужен Plus

Описание вакансии

Текст:
/
TL;DR
Data Scientist (NLP/Search) (NLP, Information Retrieval, E-commerce): Building multilingual query-understanding, semantic retrieval, and vector-search models for product discovery across Trendyol markets with an accent on Arabic-language search, spelling correction, and language-specific NLP challenges. Focus on designing evaluation methodologies, validating improvements through A/B experiments, and deploying reliable production ML systems with strong quality and latency.

Location: Istanbul / Maslak, Turkey; hybrid workplace

Company

hirify.global Tech enables e-commerce through technology and builds search and language-intelligence solutions for customers across multiple markets.

What you will do

  • Develop multilingual query-understanding capabilities, including spelling correction, normalization, rewriting, intent detection, and semantic retrieval.
  • Build and evaluate vector-search, embedding, and retrieval models to improve product-discovery relevance.
  • Lead Arabic-language search initiatives addressing dialect variation, morphology, transliteration, mixed-language queries, and spelling errors.
  • Analyze search queries, product content, user interactions, and linguistic signals across large-scale structured and unstructured datasets.
  • Design offline evaluation methods and quality metrics, validate improvements through online A/B experiments, and monitor post-deployment performance.
  • Collaborate with Search, Pre-Search, Ranking, Product, and Engineering teams to deliver scalable machine-learning solutions and improve production quality, latency, reliability, and deployment cycles.

Requirements

  • Experience in applied data science and machine learning, preferably in NLP, information retrieval, search, or language modeling.
  • Professional working proficiency in Arabic, including Arabic grammar, morphology, spelling conventions, dialectal variation, and search-query behavior.
  • Experience with transformer-based language models, text embeddings, semantic search, retrieval, reranking, query understanding, or spelling correction.
  • Strong Python and SQL skills, with practical experience using machine-learning libraries such as PyTorch, TensorFlow, scikit-learn, or Hugging Face Transformers.
  • Experience with A/B experiments, offline evaluation, statistical analysis, and ambiguous product problems.
  • Excellent written and verbal English communication skills.

Nice to have

  • Native-level Arabic proficiency.
  • Experience with Arabic NLP resources, datasets, tokenization, or normalization approaches.
  • Experience with vector databases, approximate nearest-neighbor search, embedding models, or large-scale retrieval systems.
  • Experience in e-commerce, marketplaces, search, recommendation, or self-service platforms.
  • Turkish language skills.

Culture & Benefits

  • Hybrid working model with flexibility, including work-from-abroad opportunities and a summer working model.
  • Customisable FlexBenefits budget covering meals, health insurance, fuel support, and shopping credits.
  • Well-being support including doctors, psychologist and dietitian services, and HPV vaccination.
  • Annual training and conference allowance, learning-platform access, and in-person learning sessions.
  • International collaboration with colleagues across offices in Berlin, Amsterdam, Dubai, and other locations.
  • Ownership from day one, mentoring, team rituals, events, and opportunities for professional growth.

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 →