Data scientist in fintech

IDT Corporation; United States; Hybrid; USD 140,000 - 210,000; Full time

Benefits offered by this job

Remote/hybrid работа

Оплата обучения и конференций

English classes

Job summary

IDT Corporation ищет старшего Data Scientist для разработки новых ML‑моделей с нуля и их разворачивания в реальном времени. Вы будете решать задачи по мошенничеству, рискам платежей, персонализации и финансовому прогнозированию на больших транзакционных данных, возьмете на себя полный цикл разработки — от постановки задачи до мониторинга в продакшне.

Требуется 5+ лет опыта в Data Science/ML, сильное владение Python (Scikit‑learn, Pandas, NumPy) и опытом работы с PyTorch или TensorFlow,

Qualifications

• 5+ лет hands‑on опыта в Data Science/ML с доведением моделей до продакшна
• Связь на английском не ниже Upper‑Intermediate
• Глубокое понимание статистики, вероятности и валидации
• Опыт работы с большими данными и инструментами ML
• Плюсом: FinTech/опыт в платежах/anti‑fraud

Responsibilities

• Проектирование и создание новых моделей с нуля (градиентный бустинг, нейронные сети, ансамбли)
• Поддержка валидаций и предотвращение утечки данных
• Сотрудничество с инженериями данных и MLOps для разворачивания моделей в продакшне
• Организация пайплайнов тренировки, трекинг экспериментов (MLflow) и мониторинг дрейфа
• Разработка A/B тестов и оценочных метрик
• Изучение и внедрение новых подходов (LLMs/GenAI, графовые методы) и наставничество коллег

Skills

ML моделирование

Python (Scikit‑learn, Pandas, NumPy)

PyTorch или TensorFlow

MLflow/ ClearML

Статистика и валидация

Education

Бакалавр в CS/математике/ steels

Магистр или PhD приветствуются

Tools

Snowflake

Feature Stores

Graph нейронные сети

Job description

Описание: IDT is expanding its AI Lab and seeking a hands‑on, highly experienced Senior Data Scientist with a proven track record of building new ML models from scratch and deploying them into real‑time production. You will tackle complex problems across fraud prevention, payment risk, transaction scoring, personalization, and financial forecasting using large‑scale transactional datasets. You will own the full modeling lifecycle — from problem framing through to production monitoring.

Задачи:

• Design and build new models from the ground up (gradient boosting, neural networks, ensembles, anomaly detection) for anti‑fraud, risk assessment, and personalization.
• Maintain rigorous validation practices to prevent data leakage and overfitting.
• Partner with Data Engineering and MLOps to operationalize models in production (batch and real‑time/low‑latency).
• Build training, experiment tracking (MLflow), and model/data drift monitoring pipelines.
• Design A/B tests and statistical evaluations to measure model performance against key business metrics (reduced fraud losses, transaction acceptance rates, decision latency).
• Explore emerging techniques (LLMs/GenAI, graph‑based methods), establish engineering best practices, and mentor fellow team members.

Требования:

• 5+ years of hands‑on experience in Data Science / ML, with a strong focus on designing and building new ML models from scratch and delivering them into production with measurable business impact.
• B2+ (Upper‑Intermediate or higher) — fluent written and spoken communication for technical discussions and daily collaboration with an international team.
• Deep understanding of statistics, probability, and validation methodologies; practical experience handling class imbalance, model calibration, and anomaly detection.
• Advanced proficiency (Scikit‑learn, Pandas, NumPy); Hands‑on experience with PyTorch or TensorFlow; Complex queries and large‑scale data manipulation; Experiment tracking and model lifecycle management (MLflow, ClearML, etc.).
• Bachelor’s degree in Computer Science, Mathematics, Statistics, Engineering, or a related quantitative field.
• Будет плюсом:
  • Domain experience in FinTech, Payments, or Fraud Prevention: real‑time transaction scoring and low‑latency inference;
  • Experience with cloud data platforms (Snowflake), Feature Stores, and event‑driven architectures;
  • Background in time‑series forecasting, graph neural networks, Explainable AI (XAI), or applied LLM solutions;
  • Master’s degree or Ph.D. in a quantitative discipline.

Условия:

• Remote b2b or hybrid (Belarus‑Moldova) work opportunity!
• Stable job with long‑term growth perspective
• Competitive salary with annual performance review
• Really good hardware
• An exciting and challenging job with talented people around
• Continuous learning and career growth opportunities
• Compensation for professional training, seminars, and conferences
• Referral program - get rewarded for helping us grow the team with talented people
• Company-supported English classes to enhance your professional growth
• More perks for the Minsk and Chisinau office employees