В Roblox четко очерчена senior-роль по созданию базовой ML-инфраструктуры для огромной аудитории, с современными инструментами и заметным влиянием на безопасность пользователей; зарплатный диапазон пересекается с рыночным ориентиром.
Ограничения по визам
Современный стек
Понятная зона ответственности
Влияние на безопасность
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Подключить Plus
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Senior Software Engineer-ML Platform (AI): Building high-scale ML infrastructure for user-level safety, including feature stores, embeddings, risk-score serving, and real-time behavioral data pipelines with an accent on low-latency systems, data quality, and auditability. Focus on streaming and batch architecture, productionizing models, enabling ML experimentation, and designing fair, explainable systems that affect millions of users.
Location: San Mateo, CA, United States; office-based roles are onsite Tuesday, Wednesday, and Thursday, with optional presence on Monday and Friday.
Annual base salary: $196,750–$243,290 USD. Full-time employees are also eligible for equity compensation and benefits.
Company
Roblox builds a platform for creating and participating in 3D immersive digital experiences, serving tens of millions of users and a global developer community.
What you will do
Build and operate a safety user feature store that transforms violations, reports, appeals, filter activity, and engagement signals into reliable features for training and serving.
Productionize user embeddings and risk scores with low-latency online lookup and similar-user search.
Extend batch pipelines to streaming so behavioral signals reach models and safety systems within minutes.
Provide self-serve tooling for feature discovery, training-data generation, offline evaluation, and A/B experiment hooks.
Implement monitoring for feature freshness, drift, lineage, ownership, explainability, fairness review, and appeals correction.
Own end-to-end delivery from system design through migration, testing, production operations, and on-call.
Requirements
5+ years of experience building and operating production ML platforms, data infrastructure, or large-scale backend systems.
Strong experience with batch and streaming data pipelines, such as Spark, Flink, and Kafka, feeding feature stores or similar ingestion layers.
Experience designing low-latency serving systems, including service APIs, RPC/gRPC, key-value storage, caching, and high-throughput read paths.
Proficiency in Python and a statically typed backend language such as Go, Java, or a similar language.
Production operations experience with monitoring, alerting, on-call, and incident response for systems with user-facing impact.
Working understanding of the ML lifecycle, including training-data generation, offline evaluation, deployment, and drift monitoring.
Ability to design reusable platforms, work across ML, Policy, Product, and Detection teams, and build reliable, fair, and auditable systems.
Nice to have
Experience in Trust & Safety, content moderation, user detection, or other adversarial domains.
Experience with user-level modeling or user understanding.
Experience with embeddings and vector similarity search in production.
Familiarity with feature-store platforms and signal governance.
Experience building agentic workflows or tool interfaces for LLM agents, such as MCP.
Culture & Benefits
Work focuses on reducing harm, improving fairness, and supporting safer shared experiences, including for minors.
Collaboration spans ML, Policy, Product, and Detection teams.
Full-time employees are eligible for equity compensation and benefits.
Roblox provides equal employment opportunities and reasonable recruiting accommodations for qualifying disabilities or religious beliefs.
For US-based roles, Roblox may not be able to employ candidates with certain US visa-related work authorizations or support future H-1B sponsorship.
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 →