Данные и ML вакансии - в офисе

Location: Barcelona, Spain; Tokyo, Japan; Portland, OR Workplace: onsite About Welocalize  Welocalize, a Welo Global brand, serves localization teams through AI-enabled multilingual content solutions that enable enterprises to operate and scale globally. Welocalize combines AI, automation, and human expertise to support enterprises in more than 300 languages, enabling accurate, culturally aligned, and compliant multilingual content at scale.   Welocalize’s Opal Platform comprises patented technology designed to automate multilingual content and improve workflow performance for enterprises. Opal functions as an agentic system that orchestrates AI, automation, and human expertise to coordinate content workflows across the lifecycle, improving speed, scalability, and operational control for global enterprises. These solutions operate within a secure and compliant environment supported by seven ISO certifications. Welocalize is headquartered in New York with offices worldwide. welocalize.com  ROLE OVERVIEW  Welocalize is looking for a native-level Bengali speaker to join a confidential, large-scale AI and consumer technology programme for a major global technology and communications client.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how the product actually performs on a client-supplied device — a smartphone and a next-generation smart glass. If you enjoy noticing the detail other people miss, this is work you'll find satisfying.  Welocalize is looking for a native-level Bengali speaker to join a large-scale AI and consumer technology programme.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how it performs on a real device. You'll be among the first people to see this technology in Bengali, and your judgement shapes how well it works before anyone else uses it.  If you enjoy noticing the detail other people miss, this is work you'll find satisfying.    MAIN RESPONSIBILITIES  Review pre-release AI model output in Bengali and judge it on accuracy, naturalness, appropriateness, and overall quality — would a native speaker really say it this way? This includes fact-checking each response against the source material.

Location: Barcelona, Spain; Tokyo, Japan; Portland, OR Workplace: onsite About Welocalize  Welocalize, a Welo Global brand, serves localization teams through AI-enabled multilingual content solutions that enable enterprises to operate and scale globally. Welocalize combines AI, automation, and human expertise to support enterprises in more than 300 languages, enabling accurate, culturally aligned, and compliant multilingual content at scale.   Welocalize’s Opal Platform comprises patented technology designed to automate multilingual content and improve workflow performance for enterprises. Opal functions as an agentic system that orchestrates AI, automation, and human expertise to coordinate content workflows across the lifecycle, improving speed, scalability, and operational control for global enterprises. These solutions operate within a secure and compliant environment supported by seven ISO certifications. Welocalize is headquartered in New York with offices worldwide. welocalize.com  ROLE OVERVIEW  Welocalize is looking for a native-level Urdu speaker to join a confidential, large-scale AI and consumer technology programme for a major global technology and communications client.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how the product actually performs on a client-supplied device — a smartphone and a next-generation smart glass. If you enjoy noticing the detail other people miss, this is work you'll find satisfying.  Welocalize is looking for a native-level Urdu speaker to join a large-scale AI and consumer technology programme.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how it performs on a real device. You'll be among the first people to see this technology in Urdu, and your judgement shapes how well it works before anyone else uses it.  If you enjoy noticing the detail other people miss, this is work you'll find satisfying.    MAIN RESPONSIBILITIES  Review pre-release AI model output in Urdu and judge it on accuracy, naturalness, appropriateness, and overall quality — would a native speaker really say it this way? This includes fact-checking each response against the source material.

Задачи: — Разработка и внедрение ML-моделей прогнозирования спроса (временные ряды, градиентный бустинг, нейросети) на уровне SKU; — Автоматизация пайплайнов ретрайна и мониторинга качества прогнозов в продакшене с использованием MLflow, Airflow, Docker; — Интеграция внешних факторов (промо, погода, сезонность, макроэкономика) и сценарное "what-if" моделирование; — Согласование иерархических прогнозов для оптимизации запасов в цепочке поставок (дефицит ↔ перепроизводство). Требования: — Уверенное владение Python (Pandas, NumPy, Scikit-learn); — Продвинутый уровень SQL (оконные функции, CTE, оптимизация запросов); — Опыт работы с Big Data: PySpark (распределённые данные > 100 ГБ); — Знания MLOps: Airflow / Prefect, MLflow / Weights & Biases, Docker, CI/CD; — Опыт работы с облачными платформами: AWS (SageMaker, S3, Redshift) или Azure / Yandex Cloud; — Знания статистики: теория вероятностей, проверка гипотез, A/B-тестирование, байесовские методы; — Бизнес-ориентация: понимание метрик (WMAPE, Fill Rate, оборачиваемость) и их влияния на P&L; — Умение объяснять сложные вещи простым языком нетехническим менеджерам и закупщикам; — Проактивность: самостоятельный поиск гипотез и точек роста, а не просто выполнение ТЗ; — Стрессоустойчивость: работа с "грязными" данными и умение отстаивать качество прогнозов перед бизнесом. Условия: — Работа в офисе у м. Кунцевская, бизнес-центр Kutuzoff Tower; — График работы 5/2 с 09:00 до 18:00; — Официальное оформление по ТК РФ, официальная зарплата с больничными и отпусками; — Заработная плата обсуждается на собеседовании; — Корпоративные скидки на продукцию компании (алкоголь); — Подарки сотрудникам и их детям на праздники.

Location: New York; Stockholm; San Francisco Workplace: OnSite ABOUT US: AI NEEDS A NEW INFRASTRUCTURE LAYER. WE'RE BUILDING IT AT MODAL. Every era of computing brought new workloads that previous infrastructure couldn't support: mainframes, databases, and the cloud. Each time, the company that rebuilt the layer underneath defined the decade. AI is no different, except it touches everything instead of one slice, and the window to build the layer underneath it is open right now. Our customers include category-defining companies like Ramp https://modal.com/blog/how-ramp-built-a-full-context-background-coding-agent-on-modal, Cognition, DoorDash, and Suno. They rely on Modal for instant GPU access, sub-second container starts, and native storage, so it's simple to serve low-latency inference, fine-tune models, and access production-ready sandboxes at scale. We've passed $1B in annualized run rate, up 15x from a year ago. In October 2026 we raised a $750M Series D led by Accel at a $15.75B valuation. Our team includes creators of popular open-source projects (e.g.,Seaborn https://github.com/mwaskom/seaborn,Luigi https://github.com/spotify/luigi), academic researchers, international olympiad medalists, and experienced engineering and product leaders with decades of experience. THE ROLE: We’re growing our Data team and are looking for our first few key hires to build self-serve data tools and drive business strategy in the right direction. The mission of the Modal Data team is to make it easy to track company goals, make evidence-backed decisions, and prioritize the right work. We do this via: - Self-serve AI analytics tools (Hex, Snowflake) - Embedding with teams as a “data adviser,” providing strategic analysis and consulting WHAT YOU'LL WORK ON: - Contribute to building the most modern analytics stack https://modal.com/blog/analytics-stack in data today to support AI-driven self-serve analysis, key metrics tracking, and external customer reporting - Influence work on new products like LLM Inference Endpoints https://modal.com/blog/qwen3-8-2-4t-a95b-now-available-on-modal through product analytics tracking - Identify millions of dollars of cost savings and optimization across our tools and financial operations - Write data pipelines that power the operations of our business, such as our cloud compute economics or sales comp - Create foundational datasets that can be used by people and AI tools to answer questions around product use cases, financial reporting, and marketing campaigns YOU MAY BE A FIT IF: - You’re fluent in SQL and proficient in Python - You have professional experience with at least 2 of the following tools: Snowflake, dbt, dlt, Modal, Hex, PostHog - You extend your work beyond just data reporting and into action and impact - You have high attention to detail - You’re an excellent and precise communicator - You’re personable and great at building relationships NICE-TO-HAVE: - Experience with AI products, especially LLM inference and sandboxes - Experience in fin ops, fraud, sales ops, or risk, especially in the context of AI (e.g. token cost optimization) - Project management skills - Strong presence in the data community online and offline

Location: Barcelona, Spain; Tokyo, Japan; Portland, OR Workplace: onsite About Welocalize  Welocalize, a Welo Global brand, serves localization teams through AI-enabled multilingual content solutions that enable enterprises to operate and scale globally. Welocalize combines AI, automation, and human expertise to support enterprises in more than 300 languages, enabling accurate, culturally aligned, and compliant multilingual content at scale.   Welocalize’s Opal Platform comprises patented technology designed to automate multilingual content and improve workflow performance for enterprises. Opal functions as an agentic system that orchestrates AI, automation, and human expertise to coordinate content workflows across the lifecycle, improving speed, scalability, and operational control for global enterprises. These solutions operate within a secure and compliant environment supported by seven ISO certifications. Welocalize is headquartered in New York with offices worldwide. welocalize.com  ROLE OVERVIEW  Welocalize is looking for a native-level Swedish speaker to join a confidential, large-scale AI and consumer technology programme for a major global technology and communications client.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how the product actually performs on a client-supplied device — a smartphone and a next-generation smart glass. If you enjoy noticing the detail other people miss, this is work you'll find satisfying.  Welocalize is looking for a native-level Swedish speaker to join a large-scale AI and consumer technology programme.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how it performs on a real device. You'll be among the first people to see this technology in Swedish, and your judgement shapes how well it works before anyone else uses it.  If you enjoy noticing the detail other people miss, this is work you'll find satisfying.    MAIN RESPONSIBILITIES  Review pre-release AI model output in Swedish and judge it on accuracy, naturalness, appropriateness, and overall quality — would a native speaker really say it this way? This includes fact-checking each response against the source material.

Location: Barcelona, Spain; Tokyo, Japan; Portland, OR Workplace: onsite About Welocalize  Welocalize, a Welo Global brand, serves localization teams through AI-enabled multilingual content solutions that enable enterprises to operate and scale globally. Welocalize combines AI, automation, and human expertise to support enterprises in more than 300 languages, enabling accurate, culturally aligned, and compliant multilingual content at scale.   Welocalize’s Opal Platform comprises patented technology designed to automate multilingual content and improve workflow performance for enterprises. Opal functions as an agentic system that orchestrates AI, automation, and human expertise to coordinate content workflows across the lifecycle, improving speed, scalability, and operational control for global enterprises. These solutions operate within a secure and compliant environment supported by seven ISO certifications. Welocalize is headquartered in New York with offices worldwide. welocalize.com  ROLE OVERVIEW  Welocalize is looking for a native-level Finnish speaker to join a confidential, large-scale AI and consumer technology programme for a major global technology and communications client.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how the product actually performs on a client-supplied device — a smartphone and a next-generation smart glass. If you enjoy noticing the detail other people miss, this is work you'll find satisfying.  Welocalize is looking for a native-level Finnish speaker to join a large-scale AI and consumer technology programme.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how it performs on a real device. You'll be among the first people to see this technology in Finnish, and your judgement shapes how well it works before anyone else uses it.  If you enjoy noticing the detail other people miss, this is work you'll find satisfying.    MAIN RESPONSIBILITIES  Review pre-release AI model output in Finnish and judge it on accuracy, naturalness, appropriateness, and overall quality — would a native speaker really say it this way? This includes fact-checking each response against the source material.

Location: Barcelona, Spain; Tokyo, Japan; Portland, OR Workplace: onsite About Welocalize  Welocalize, a Welo Global brand, serves localization teams through AI-enabled multilingual content solutions that enable enterprises to operate and scale globally. Welocalize combines AI, automation, and human expertise to support enterprises in more than 300 languages, enabling accurate, culturally aligned, and compliant multilingual content at scale.   Welocalize’s Opal Platform comprises patented technology designed to automate multilingual content and improve workflow performance for enterprises. Opal functions as an agentic system that orchestrates AI, automation, and human expertise to coordinate content workflows across the lifecycle, improving speed, scalability, and operational control for global enterprises. These solutions operate within a secure and compliant environment supported by seven ISO certifications. Welocalize is headquartered in New York with offices worldwide. welocalize.com  ROLE OVERVIEW  Welocalize is looking for a native-level Hindi speaker to join a confidential, large-scale AI and consumer technology programme for a major global technology and communications client.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how the product actually performs on a client-supplied device — a smartphone and a next-generation smart glass. If you enjoy noticing the detail other people miss, this is work you'll find satisfying.  Welocalize is looking for a native-level Hindi speaker to join a large-scale AI and consumer technology programme.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how it performs on a real device. You'll be among the first people to see this technology in Hindi, and your judgement shapes how well it works before anyone else uses it.  If you enjoy noticing the detail other people miss, this is work you'll find satisfying.    MAIN RESPONSIBILITIES  Review pre-release AI model output in Hindi and judge it on accuracy, naturalness, appropriateness, and overall quality — would a native speaker really say it this way? This includes fact-checking each response against the source material.

Location: Barcelona, Spain; Tokyo, Japan; Portland, OR Workplace: onsite About Welocalize  Welocalize, a Welo Global brand, serves localization teams through AI-enabled multilingual content solutions that enable enterprises to operate and scale globally. Welocalize combines AI, automation, and human expertise to support enterprises in more than 300 languages, enabling accurate, culturally aligned, and compliant multilingual content at scale.   Welocalize’s Opal Platform comprises patented technology designed to automate multilingual content and improve workflow performance for enterprises. Opal functions as an agentic system that orchestrates AI, automation, and human expertise to coordinate content workflows across the lifecycle, improving speed, scalability, and operational control for global enterprises. These solutions operate within a secure and compliant environment supported by seven ISO certifications. Welocalize is headquartered in New York with offices worldwide. welocalize.com  ROLE OVERVIEW  Welocalize is looking for a native-level Polish speaker to join a confidential, large-scale AI and consumer technology programme for a major global technology and communications client.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how the product actually performs on a client-supplied device — a smartphone and a next-generation smart glass. If you enjoy noticing the detail other people miss, this is work you'll find satisfying.  Welocalize is looking for a native-level Polish speaker to join a large-scale AI and consumer technology programme.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how it performs on a real device. You'll be among the first people to see this technology in Polish, and your judgement shapes how well it works before anyone else uses it.  If you enjoy noticing the detail other people miss, this is work you'll find satisfying.    MAIN RESPONSIBILITIES  Review pre-release AI model output in Polish and judge it on accuracy, naturalness, appropriateness, and overall quality — would a native speaker really say it this way? This includes fact-checking each response against the source material.

Location: Barcelona, Spain; Tokyo, Japan; Portland, OR Workplace: onsite About Welocalize  Welocalize, a Welo Global brand, serves localization teams through AI-enabled multilingual content solutions that enable enterprises to operate and scale globally. Welocalize combines AI, automation, and human expertise to support enterprises in more than 300 languages, enabling accurate, culturally aligned, and compliant multilingual content at scale.   Welocalize’s Opal Platform comprises patented technology designed to automate multilingual content and improve workflow performance for enterprises. Opal functions as an agentic system that orchestrates AI, automation, and human expertise to coordinate content workflows across the lifecycle, improving speed, scalability, and operational control for global enterprises. These solutions operate within a secure and compliant environment supported by seven ISO certifications. Welocalize is headquartered in New York with offices worldwide. welocalize.com  ROLE OVERVIEW  Welocalize is looking for a native-level Persian speaker to join a confidential, large-scale AI and consumer technology programme for a major global technology and communications client.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how the product actually performs on a client-supplied device — a smartphone and a next-generation smart glass. If you enjoy noticing the detail other people miss, this is work you'll find satisfying.  Welocalize is looking for a native-level Persian speaker to join a large-scale AI and consumer technology programme.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how it performs on a real device. You'll be among the first people to see this technology in Persian, and your judgement shapes how well it works before anyone else uses it.  If you enjoy noticing the detail other people miss, this is work you'll find satisfying.    MAIN RESPONSIBILITIES  Review pre-release AI model output in Persian and judge it on accuracy, naturalness, appropriateness, and overall quality — would a native speaker really say it this way? This includes fact-checking each response against the source material.

Location: Barcelona, Spain; Tokyo, Japan; Portland, OR Workplace: onsite About Welocalize  Welocalize, a Welo Global brand, serves localization teams through AI-enabled multilingual content solutions that enable enterprises to operate and scale globally. Welocalize combines AI, automation, and human expertise to support enterprises in more than 300 languages, enabling accurate, culturally aligned, and compliant multilingual content at scale.   Welocalize’s Opal Platform comprises patented technology designed to automate multilingual content and improve workflow performance for enterprises. Opal functions as an agentic system that orchestrates AI, automation, and human expertise to coordinate content workflows across the lifecycle, improving speed, scalability, and operational control for global enterprises. These solutions operate within a secure and compliant environment supported by seven ISO certifications. Welocalize is headquartered in New York with offices worldwide. welocalize.com  ROLE OVERVIEW  Welocalize is looking for a native-level Portuguese Brazilian speaker to join a confidential, large-scale AI and consumer technology programme for a major global technology and communications client.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how the product actually performs on a client-supplied device — a smartphone and a next-generation smart glass. If you enjoy noticing the detail other people miss, this is work you'll find satisfying.  Welocalize is looking for a native-level Portuguese Brazilian speaker to join a large-scale AI and consumer technology programme.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how it performs on a real device. You'll be among the first people to see this technology in Portuguese Brazilian , and your judgement shapes how well it works before anyone else uses it.  If you enjoy noticing the detail other people miss, this is work you'll find satisfying.    MAIN RESPONSIBILITIES  Review pre-release AI model output in Portuguese Brazilian and judge it on accuracy, naturalness, appropriateness, and overall quality — would a native speaker really say it this way? This includes fact-checking each response against the source material.

Location: Barcelona, Spain; Tokyo, Japan; Portland, OR Workplace: onsite About Welocalize  Welocalize, a Welo Global brand, serves localization teams through AI-enabled multilingual content solutions that enable enterprises to operate and scale globally. Welocalize combines AI, automation, and human expertise to support enterprises in more than 300 languages, enabling accurate, culturally aligned, and compliant multilingual content at scale.   Welocalize’s Opal Platform comprises patented technology designed to automate multilingual content and improve workflow performance for enterprises. Opal functions as an agentic system that orchestrates AI, automation, and human expertise to coordinate content workflows across the lifecycle, improving speed, scalability, and operational control for global enterprises. These solutions operate within a secure and compliant environment supported by seven ISO certifications. Welocalize is headquartered in New York with offices worldwide. welocalize.com  ROLE OVERVIEW  Welocalize is looking for a native-level Arabic speaker to join a confidential, large-scale AI and consumer technology programme for a major global technology and communications client.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how the product actually performs on a client-supplied device — a smartphone and a next-generation smart glass. If you enjoy noticing the detail other people miss, this is work you'll find satisfying.  Welocalize is looking for a native-level Arabic speaker to join a large-scale AI and consumer technology programme.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how it performs on a real device. You'll be among the first people to see this technology in Arabic, and your judgement shapes how well it works before anyone else uses it.  If you enjoy noticing the detail other people miss, this is work you'll find satisfying.    MAIN RESPONSIBILITIES  Review pre-release AI model output in Arabic and judge it on accuracy, naturalness, appropriateness, and overall quality — would a native speaker really say it this way? This includes fact-checking each response against the source material.

Location: Barcelona, Spain; Tokyo, Japan; Portland, OR Workplace: onsite About Welocalize  Welocalize, a Welo Global brand, serves localization teams through AI-enabled multilingual content solutions that enable enterprises to operate and scale globally. Welocalize combines AI, automation, and human expertise to support enterprises in more than 300 languages, enabling accurate, culturally aligned, and compliant multilingual content at scale.   Welocalize’s Opal Platform comprises patented technology designed to automate multilingual content and improve workflow performance for enterprises. Opal functions as an agentic system that orchestrates AI, automation, and human expertise to coordinate content workflows across the lifecycle, improving speed, scalability, and operational control for global enterprises. These solutions operate within a secure and compliant environment supported by seven ISO certifications. Welocalize is headquartered in New York with offices worldwide. welocalize.com  ROLE OVERVIEW  Welocalize is looking for a native-level Turkish speaker to join a confidential, large-scale AI and consumer technology programme for a major global technology and communications client.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how the product actually performs on a client-supplied device — a smartphone and a next-generation smart glass. If you enjoy noticing the detail other people miss, this is work you'll find satisfying.  Welocalize is looking for a native-level Turkish speaker to join a large-scale AI and consumer technology programme.  It's a hands-on, on-site role that combines two things: evaluating a pre-release AI model in your language, and testing how it performs on a real device. You'll be among the first people to see this technology in Turkish, and your judgment shapes how well it works before anyone else uses it.  If you enjoy noticing the detail other people miss, this is work you'll find satisfying.    MAIN RESPONSIBILITIES  Review pre-release AI model output in Turkish and judge it on accuracy, naturalness, appropriateness, and overall quality — would a native speaker really say it this way? This includes fact-checking each response against the source material.

Задачи: — Проектирование и развитие процессов сбора, загрузки, преобразования и передачи данных (ETL/ELT) — Разработка и сопровождение пайплайнов обработки больших объемов данных — Работа с реляционными и нереляционными базами данных — Анализ структуры и качества данных, выявление дублирования, пропусков и несоответствий — Разработка SQL-запросов, процедур, скриптов и механизмов обработки данных — Построение интеграционного взаимодействия между информационными системами и источниками данных — Участие в разработке и развитии DWH / Data Lake / аналитических хранилищ — Настройка процессов регулярной загрузки и синхронизации данных — Обеспечение контроля качества, полноты и целостности данных — Оптимизация запросов и процессов обработки больших объемов данных — Анализ источников данных и определение необходимых таблиц, полей, связей и бизнес-смыслов — Участие в проектировании моделей данных и структуры хранилищ — Взаимодействие с системными и бизнес-аналитиками, разработчиками, архитекторами и DevOps — Подготовка технической документации по потокам данных, интеграциям и структуре БД — Участие в тестировании и приемке интеграционных решений — Поиск и устранение проблем в процессах загрузки и обработки данных — Участие в реализации интеграционных решений между крупными государственными информационными системами Требования: — Опыт работы от 3 лет в роли Data Engineer / инженера по данным — Опыт работы с большими объемами данных и высоконагруженными системами — Уверенное знание SQL и опыт работы с СУБД — Понимание принципов построения DWH, Data Lake и ETL/ELT — Опыт разработки и сопровождения пайплайнов обработки данных — Навыки анализа структуры, качества и взаимосвязей данных — Опыт работы с реляционными и нереляционными базами данных — Понимание принципов интеграции информационных систем — Готовность работать с крупными государственными и медицинскими информационными системами Условия: — Работа над масштабным государственным проектом в сфере цифрового здравоохранения Казахстана — Работа с большими объемами данных и сложными интеграциями — Участие в развитии архитектуры данных и хранилища — Профессиональный рост и развитие компетенций Data Engineering — Взаимодействие с командой аналитиков, разработчиков, архитекторов и DevOps — Оформление по ГПХ / ДВОУ — График 5/2, 09:00–18:00 — Обед 13:00–14:00 — Формат работы: только офис — Адрес: г. Астана, пр. Кабанбай батыра, 17

Задачи: — Разработка и развитие сервисов промышленного инференса ML-моделей и LLM — Построение централизованного слоя доступа к моделям и AI-сервисам — Разработка платформенных компонентов для деплоя, версионирования и мониторинга моделей

Команда SberNBA занимается тремя большими направлениями: платформа коммуникаций и персональных предложений Сбера, поведенческая foundation model человека, бизнес направление коммуникаций от потребности клиента. Цель команды – сделать управление коммуникациями проще и эффективнее за счет гибкости платформенных инструментов и GenAI, моделировать поведение человека на краткосрочном и долгосрочном горизонте и использовать эти данные в бизнес-процессах банка, помогать человеку закрывать потребности быстрее и проще. Обязанности: — развивать команду и компетенции трайба на уровне DS экспертизы; — строить, развивать и реализовывать стратегию развития трайба в части ML/AI; — участвовать в создании кампейн платформы нового поколения, основанного на агентах; — развивать системы генерации персонализированного контента; — развивать Foundation Model (трансформерная модель понимания клиента на основе его действий в банке и экосистеме Сбера) и рекомендательные модели. Требования: — 10+ лет в DS / 5+ лет в промышленных RecSys в крупных компаниях; — интеграция LLM в RecSys (RAG / генерация текста офферов); — опыт управления командой 10+ DS и развития ML экспертизы на уровне подразделения; — успешный опыт реализации сложных и масштабных ML проектов с доказанным бизнес-эффектом; — хорошие коммуникативные навыки, готовность договариваться и находить оптимальные решения; — стратегическое мышление; — амбициозность и ориентация на бизнес-результат. Условия: — комфортный современный офис рядом с м. Кутузовская; — ежегодный пересмотр зарплаты и годовая премия; — корпоративный спортзал и зоны отдыха; — более 400 образовательных программ СберУниверситета для профессионального и карьерного развития; — расширенный ДМС, льготное страхование для семьи и корпоративная пенсионная программа; — гибкий дисконт по ипотечному кредиту, равный 1/3 ключевой ставки ЦБ; — подписка Прайм с возможностью совместного использования на трёх близких; — вознаграждение за рекомендацию друзей в команду Сбера.

Задачи: — глубокий анализ данных для построения и калибровки моделей — опыт построения uplift-моделей и кривых отклика для оптимизации процентных ставок — разработка моделей оценки эластичности спроса на кредит по ставке — разработка, обучение и валидация продвинутых ML-моделей для индивидуального ценообразования (регрессионные модели, градиентный бустинг, методы обучения с подкреплением / Reinforcement Learning для динамического ценообразования) — анализ и интерпретация результатов A/B-тестов ценовых стратегий, оценка их бизнес-эффекта (влияние на выручку, маржу, конверсию, LTV) — построение и оптимизация симуляционных сред для предварительной оценки эффекта ценовых изменений. Требования: — глубокое знание и практический опыт (от 3 лет) применения методов машинного обучения и анализа данных для решения бизнес-задач, желательно в области ценообразования, экономики или финтеха — продвинутый уровень знаний в области статистики, эконометрики и методологии проведения A/B-тестов — понимание принципов микроэкономики и теории цен — опыт разработки и внедрения в production моделей, напрямую влияющих на финансовые метрики (выручка, маржа) — продвинутые знания SQL и Python для анализа данных и ML — опыт работы с Python ML-фреймворками (Scikit-Learn, LightGBM/CatBoost/XGBoost, PyTorch/TensorFlow, библиотеки для RL), Spark (PySpark), опыт работы с базами данных (GreenPlum, PostgreSQL, ClickHouse). Будет плюсом: — опыт построения систем динамического ценообразования (dynamic pricing) или моделей на основе обучения с подкреплением (Reinforcement Learning) — опыт работы в ритейле, финтехе, marketplace-сегменте или сфере услуг с высокой конкуренцией — умение переводить бизнес-задачи на язык данных и обратно — знание основ юнит-экономики и финансового моделирования — опыт разработки или интеграции решений с применением GenAI. Условия: — комфортный современный офис рядом с м. Кутузовская — офисный формат работы — ежегодный пересмотр зарплаты, квартальная и годовая премия — корпоративный спортзал и зоны отдыха — более 400 образовательных программ СберУниверситета для профессионального и карьерного развития — программа адаптации и помощь руководителя на старте — расширенный ДМС, льготное страхование для семьи и корпоративная пенсионная программа — гибкий дисконт по ипотечному кредиту, равный 1/3 ключевой ставки ЦБ — подписка Прайм с возможностью совместного использования на трёх близких — вознаграждение за рекомендацию друзей в команду Сбера.

В данной роли предстоит заниматься разработкой и проектированием AI-агентов, которые автоматизируют процессы отчетности и выявления отклонений в бизнес-процессах. Задачи включают создание пайплайнов для обнаружения аномалий, интеграцию этих решений с большими языковыми моделями (LLM) для генерации текстовых отчетов, а также работу с большими данными: подготовку признаков, агрегацию и передачу данных в модели. Кроме того, необходимо оптимизировать, упаковывать и разворачивать агентные системы в продуктивной среде. Важной частью работы является тесное взаимодействие с бизнес-подразделениями и аналитиками для преобразования их требований в технические AI-решения. Кандидат должен иметь не менее трех лет опыта коммерческой разработки на Python, хорошо разбираться в языковых моделях, включая понимание их ограничений, а также иметь опыт работы с LangChain или аналогичными инструментами для интеграции LLM в бизнес-задачи. Требуется знание математической статистики и основ различных методов машинного обучения, а также понимание принципов DevOps и процессов CI/CD. Работа предполагает офисный формат в Москве.