4 дня назад
Data Scientist (AI)
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Описание вакансии
Текст:
TL;DR
Data Scientist (AI): Analyzing complex datasets to detect synthetic identities and fraud patterns with an accent on signal discovery and model validation. Focus on building internal tools, shaping product strategy, and ensuring the integrity of the trust layer in the AI era.
Company
is a security-focused startup building the trust layer for the internet to combat AI-driven identity fraud and deepfakes.
What you will do
- Master internal data sources to understand how identity signals are generated.
- Explore raw datasets to identify anomalies and fraud patterns.
- Build analyses to guide product strategy and model development.
- Develop internal tools and scripts using Python and SQL to validate data assumptions.
- Design analytics dashboards to provide transparency into the trust layer for customers.
- Collaborate with engineering to ensure high-quality instrumentation and model performance.
Requirements
- Highly skilled in SQL and Python.
- Proficiency with Elasticsearch, KQL, and DSL query languages.
- Strong intuition for working with messy, real-world data.
- Detail-oriented mindset with a focus on data quality and correctness.
- Ability to thrive in ambiguous environments without pre-defined roadmaps.
- Collaborative approach to working across product, engineering, and ML teams.
Nice to have
- Experience with Django.
- Background in fraud, trust & safety, or risk systems.
- Experience building customer-facing analytics.
- Familiarity with ML workflows or model evaluation.
Culture & Benefits
- Competitive salary and meaningful equity package.
- Comprehensive health benefits.
- 3 weeks of paid vacation.
- Provision of a new laptop and necessary gear.
- Opportunity to work directly with leadership on a high-impact, generational company.
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 →