Вакансии

Location: Europe; United States Workplace: Remote ABOUT CANTINA Cantina Labs is a social AI company, developing a suite of advanced real-time models that push the boundaries of expression, personality, and realism. We bring characters to life, transforming how people tell stories, connect, and create. We build and power ecosystems. Cantina, our flagship social AI platform, is just the beginning. If you're excited about the potential AI has to shape human creativity and social interactions, join us in building the future! ABOUT THE ROLE We’re working on real-time video generation for interactive characters, spanning realistic humans, stylized characters, and non-human identities. We want these characters to feel present: to speak, react, and express themselves through their face, gestures, and body language. This brings together a set of closely connected research problems. A character needs a consistent appearance, expressive and coherent motion, and a natural relationship between speech and movement. The model also needs to respond quickly and keep generating reliably as an interaction unfolds. Quality, controllability, and efficiency all matter. You’ll work across streaming generation, fine-tuning, distillation, inference optimization, and data. Some problems will call for a new training approach; others will come down to better examples, a more useful evaluation, or understanding a bottleneck in the implementation. Part of the job is figuring out which intervention is worth making. You’ll have substantial freedom to propose directions, investigate unfamiliar ideas, and shape what we work on next. We’re looking for someone who enjoys that freedom and can turn it into a steady sequence of useful experiments. WHAT YOU’LL WORK ON - Real-time and streaming generation. Develop models that support responsive interaction while maintaining visual quality and consistency over time. Investigate latency, continuity, and how model behavior changes during longer generation. - Character expression and motion. Explore how appearance, facial expression, gesture, and body movement come together into a coherent character. Improve the relationship between speech and motion across different identities and visual styles. - Fine-tuning and post-training. Adapt video models to new data, behaviors, and capabilities. Investigate training objectives and conditioning methods, and follow promising ideas through implementation and evaluation. - Distillation and inference optimization. Make generation faster and more efficient. Work across model-level approaches and runtime improvements, measuring their effects on quality, latency, and memory. - Data development. Spend meaningful time with the actual training examples. Identify missing coverage and recurring failure modes, improve curation and sampling, and develop data that helps models learn the behavior we want. - Evaluation. Build comparisons that make progress visible and failures understandable. Combine quantitative measurements with careful inspection of generated video and the experience of interacting with the model. - Research and collaboration. Read relevant work, reproduce and adapt useful ideas, and share early experiments, demos, and findings. Help the team decide what deserves more time and compute. WHAT YOU’LL BRING - Experience training or fine-tuning video or related generative models. - Ability to learn unfamiliar methods quickly and turn ideas into useful experiments. - Good experimental judgment, ownership, and clear communication. You’ll have freedom to explore new approaches. We expect fast iteration and early sharing of results. COMPENSATION When determining compensation, a number of factors will be considered, including skills, experience, job scope, location, and competitive compensation market data.   Benefits for U.S.-based roles: - Competitive salary and generous company equity - Medical, dental, and vision insurance – 99.99% of premiums covered by Cantina - 42 days of paid time off, including: - 15 PTO days - 10 sick days - 15 company holidays - 2 floating holidays - Generous parental leave & fertility support - 401(k) retirement savings plan - Lifestyle spending account – $500/month to use however you’d like - Complimentary lunch and snacks for in-office employees - One Medical membership, and more!

BlueTuskr is a full-service e-commerce growth agency focused on paid media, owned media, creative, and analytics to help clients scale profitably. The Opportunity: We're looking for a Graphic Designer and Video Editor who creates with conversion in mind. This hands-on role involves designing for e-commerce brands including emails, landing pages, website sections, and social/ad creative. The candidate will work closely with email, CRO, Shopify development, and paid media teams, and use Claude and other AI tools to enhance workflow efficiency. What You'll Own: — Design campaign and automated flow emails (newsletters, welcome, abandoned cart, post-purchase, win-back, etc.) that are on-brand and built to drive clicks and revenue — Design mobile-first emails with clear hierarchy, scannable layouts, and strong calls to action — Prepare email-ready files and specs for clean builds in Klaviyo and other email platforms — Design landing pages, homepage, collection page sections, product page modules, banners, and promotional assets for Shopify stores — Partner with CRO team to design test variations based on data, heatmaps, and hypotheses — Hand off dev-ready Figma files to web developers with desktop and mobile layouts, organized layers, and clear specs — Design static social and paid ad creative for Meta, TikTok, and other channels — Create ad variations and iterations to support creative testing — Video editing for social and ad use (resizing, trimming, captions, simple motion) where needed — Follow client brand guidelines closely and flag inconsistencies — Apply design and e-commerce best practices including accessibility, contrast, legibility, load-friendly assets, and platform specs — Present work with clear rationale to internal teams and clients — Take feedback well and manage multiple accounts and deadlines — Use Claude and other AI tools to speed up concepting, copy-to-layout exploration, asset variations, and repetitive production tasks while maintaining confidentiality and AI use guidelines Who You Are: — Exceptional designer with strong eye for layout, typography, and color, supported by a portfolio — Conversion-minded, focusing on customer needs and actions — Detail-obsessed, catching misalignments, off-brand colors, and typos — Clear communicator who explains design choices and keeps team updated — Excellent time manager able to juggle multiple clients and deadlines remotely and asynchronously — AI-curious and capable, experienced with Claude or similar AI tools — Adaptable to different brand styles What Success Looks Like in First 90 Days: — Learned brand guidelines and design systems for assigned clients — Consistently deliver email, web, and social designs on time with minimal revisions — Handed off at least one web project or CRO test variation to Shopify dev team with clean, dev-ready files — Integrated AI into workflow to speed production — Earned trust of internal teams who come with ideas Must-Have: — 2–4 years of e-commerce digital design experience (email, website, social, etc.) — Portfolio showing e-commerce work with context on goals and results — Advanced proficiency in Figma — Proficiency in Adobe Creative Suite (Photoshop, Illustrator, InDesign) — Hands-on experience using Claude and/or other AI tools in design workflow — Experience designing for web and supporting website creative (landing pages, site sections, PDPs) — Strong understanding of conversion-focused design and e-commerce best practices — Proven ability to follow and apply brand guidelines across multiple brands — Strong written and spoken English for async communication with U.S.-based team Strong Differentiators: — Video editing skills (Premiere Pro, After Effects, CapCut, or similar) — Experience designing for Klaviyo or other email platforms — Familiarity with Shopify themes and how designs get built — Experience designing A/B test variations or working with CRO data — Paid social ad creative experience (Meta, TikTok) — Agency background preferred This role is fully remote. Time Commitment and Availability: — Starts part-time at 15 hours/week with opportunity to scale — Availability between 9 a.m. and 6 p.m. EST Compensation: — $20–25 USD/hour based on experience and portfolio — Part-time, remote contractor role — Video editing experience is a strong plus and factored into the rate.

Задачи: • Придумывать и писать короткие сценарии с интригой, эмоциональным и сексуальным напряжением, сюжетными поворотами и клиффхэнгерами. • Генерировать вертикальные видео с помощью AI-инструментов и собирать из них цельные серии. • Сохранять узнаваемость персонажей от кадра к кадру и от серии к серии: внешность, волосы, одежду, голос и манеру поведения. • Создавать английскую озвучку и диалоги. • Выстраивать сюжет на несколько серий с одними и теми же героями. Требования: • Опыт работы с AI-видео: Kling, Hailuo, Seedance, Runway, Pika, Veo, Luma или аналогами. • Умение сохранять персонажей и визуальный стиль между кадрами и сериями. • Навыки монтажа и саунд-дизайна: CapCut, Premiere Pro, DaVinci Resolve или аналоги. • Понимание принципов удержания внимания в коротких вертикальных видео. • Готовность работать с откровенным эротическим контентом для взрослой аудитории. • Английский на уровне написания и редактирования коротких естественных диалогов. Будет плюсом: • Опыт создания эротического AI-контента для взрослых платформ. • Опыт создания мини-сериалов из 5–15 эпизодов с одними и теми же персонажами. Условия: • Оплата $1 000–3 000 в месяц, в зависимости от опыта креатора и объёма работы, который он может выполнять. Что прислать при отклике: • Резюме | профиль или сопроводительное письмо с указанием ФИО, описанием опыта работы с AI-видео и реализованных проектов, а также дат и периодов работы над ними. • Портфолио и 2–5 примеров AI-видео со ссылками на Google Drive. • Краткое описание инструментов, использованных при создании видео.

VK Видео Live — это стриминговая часть платформы VK Видео, которая полностью отвечает за прямые эфиры, интерактив и real-time коммуникацию с пользователями. Наша команда создаёт и развивает собственный продукт полного цикла — от запуска трансляций и механик монетизации (донаты, платные подписки, VK Donut) до аналитических кабинетов авторов. Кроме того, мы разрабатываем Live SDK — набор компонентов для встраивания чатов и управления стримами в другие продукты компании, включая крупнейшие интеграции с VK Видео и ВКонтакте. Задачи — Разрабатывать real-time интерактив для зрителей: технологичные чаты с модерацией, тредами и ботами, а также умные системы рекомендаций и уведомлений — Создавать инструменты монетизации и удержания: развивать бизнес-логику платформы (управление подписками, каналами и пользовательскими настройками) — Запускать новые фичи для авторов контента: проектировать инструменты проведения эфиров, механики рейдов, интерактивные прогнозы и модули глубокой аналитики — Проектировать высоконагруженные real-time системы: уходить от классического CRUD в сторону сложной backend-архитектуры — Писать сервисы для работы с потоковыми данными и механиками распределения метаданных — Оптимизировать сетевое взаимодействие: погружаться в низкоуровневую работу с сетевыми протоколами для минимизации задержек Требования — Уверенное знание языка Go и понимание его тонкостей — Опыт реализации и проектирования приложений на Go — Опыт разработки REST-сервисов и GRPC-сервисов на Go — Умение пользоваться отладчиком и профайлерами Go — Опыт работы с БД Postgres — Опыт работы с NoSQL БД — Опыт работы с Kafka или иными брокерами сообщений — Уверенное владение Git — Уверенный опыт работы с Linux-системами — Понимание принципов работы сетей и сетевых протоколов Будет плюсом — Понимание и опыт имплементации кодогенераторов — Опыт работы с ClickHouse — Опыт работы с Centrifugo/WebSockets — Опыт работы с K8s — Опыт конфигурации GitLab CI — Опыт работы с S3 — Понимание принципов безопасности информационных систем — Понимание принципов отказоустойчивости и масштабируемости информационных систем

Требования: — 6+ месяцев опыта с методами верификации дизайна.

Задачи: — Редактировать короткие видео (Reels, TikToks, Shorts) — Разрабатывать креативные концепции, генерировать идеи и координировать утверждение контента — Управлять и развивать аккаунты в социальных сетях — Планировать и реализовывать контент-календари — Отслеживать эффективность контента и выявлять новые тренды Требования: — Опыт в игровой индустрии или страсть к играм (PlayStation, Xbox, PC) — Опыт видеомонтажа (CapCut, Premiere Pro или аналогичные программы) — Креативное мышление и понимание трендов социальных сетей — Опыт управления аккаунтами в соцсетях — Умение работать самостоятельно и соблюдать сроки — Хорошие коммуникативные и организационные навыки Условия: — Работа в креативной и динамичной команде — Свобода экспериментов с новыми идеями — Возможности профессионального роста — Конкурентная оплата труда

HeyGen ищет инженера по прикладному искусственному интеллекту для работы с видеоагентами, который будет заниматься разработкой и оценкой функций AI-агентов в различных продуктах компании. В обязанности входит создание и тестирование новых возможностей с использованием технологий и методов, таких как feature flag, prompt engineering, A/B тестирование. Требуется опыт программирования на Go, Python, TypeScript, а также навыки работы с React и full stack разработкой. Кандидат должен уметь проводить оценку эффективности AI-агентов и интегрировать их в продуктовую экосистему. Зарплата варьируется от 180,000 до 240,000 долларов.

Location: Philippines; Indonesia Paired is a global staffing and recruiting agency that specializes in pairing remote work with top-tier talent. We admire the importance of innovative social media strategies and are committed to connecting talented individuals with great companies that need their unique skills. Our mission is to provide great jobs to talented people, no matter where they are located. We’re looking for an AI Video Ad Creator to produce short, engaging video ads for a growing D2C sports nutrition brand focused on runners and active consumers. You’ll take scripts written by the creative strategist and turn them into polished AI-generated video ads , experimenting with different visual styles including animated, stylized, and photorealistic content. Each video should ultimately showcase the brand’s real product. What You’ll Do Create 1–2 finished AI-generated videos per day , typically 15–60 seconds long. Turn creative scripts into compelling visual concepts using AI video and image generation tools. Write and refine prompts to achieve the desired look, movement, and storytelling. Generate multiple takes and variations, then select the strongest outputs. Experiment with different styles, including animated, stylized, and photorealistic visuals. Incorporate the brand’s real product into the final video. Handle light editing, captions, and final polishing. Iterate on generations until the footage meets the required creative and quality standards. Work closely with the creative strategist to bring scripts and concepts to life. AI Tools Experience with current AI video and image generation tools is essential. The team currently works with tools such as: Higgsfield Kling Veo Other current AI video/image generation platforms Requirements Hands-on experience creating AI-generated video and image content . A portfolio demonstrating your AI video work. Strong understanding of prompting and AI generation workflows. Patience and attention to detail, with a willingness to iterate until the output looks right. Ability to work across different visual styles and creative concepts. Strong creative judgment when selecting the best generations and takes. Ability to consistently deliver 1–2 finished videos per day . Nice to Have Video editing experience. Motion design experience. Strong visual/design sense. Experience creating paid social or performance marketing ads . Experience working with D2C or e-commerce brands. About the Brand The client is a growing D2C sports nutrition brand that makes energy gels using organic Yucatan honey, electrolytes, and green tea caffeine. Its products are designed for runners and active consumers, with a strong presence in the running and fitness community. The brand sells online through e-commerce and subscription channels, with additional wholesale distribution. It is a fast-growing company where creative content plays an important role in customer acquisition and brand growth. What We’re Looking For We’re looking for someone who is creative, hands-on, and highly iterative —someone who enjoys experimenting with AI tools and knows that getting a great result often takes multiple generations and refinements. You should be comfortable working independently while collaborating closely with a creative strategist to turn scripts into compelling visual ads. Benefits Competitive salary. Ability to work remotely. Significant room for growth over time

Задачи: — Создавать и адаптировать видео-креативы для performance-рекламы. — Работать с Claude/AI-агентами: ставить задачи, оценивать результат, давать правки. — Следить за качеством креативов: хук, темп, визуальная подача, читаемость, эмоция. — Быстро пробовать новые AI-подходы и внедрять то, что ускоряет production. — При необходимости создавать или дорабатывать руками, если так быстрее или лучше. Требования: — Опыт в motion design, video editing или рекламных видео. — Сильная насмотренность и понимание, почему креатив цепляет или не цепляет. — Умение формулировать задачи и правки: для AI, для себя и для команды. — Интерес к AI-first production и желание уходить от ручного монтажа в сторону работы с AI/агентами. — Скорость, самостоятельность, готовность много тестировать и переделывать. — Портфолио / showreel / примеры видео обязательны. Плюсом: — Опыт с performance-креативами, UGC, mobile apps или subscription-продуктами. — Понимание креативных тестов в Meta / TikTok / Snapchat. — Опыт работы с Claude/ChatGPT или другими AI-инструментами в реальных задачах. Условия: — Full-time / long-term. — Remote. — Работа на зарубежные рынки. — Кандидатов из РФ не рассматриваем.

Responsibilities: — Develop and adapt computer vision/video analytics models for detection, tracking, and recognition. — Work with video pipelines, including RTSP/HTTP, decoding, sampling, and batching. — Fine-tune models for real-world conditions and customer scenarios. — Optimize CPU inference, latency, and resource usage. — Integrate models into backend, edge, and on-premise solutions. — Prepare datasets, build training/testing pipelines, and monitor quality. Requirements: — 4+ years in Computer Vision/Video Analytics. — Strong CNN, YOLO, and SSD knowledge. — PyTorch or TensorFlow, OpenCV, and ONNX. — Understanding of RTSP, H.264/H.265, FPS, bitrate, and latency. — CPU inference optimization experience. — Linux and OpenVINO. — C++ for performance-critical components.

Задачи: — Делать AI-ролики (1 мин), в основном UGC: реалистичные персонажи, говорящие в камеру, липсинк, сюжет и сильный хук в первые секунды — Адаптировать креативы под разные гео и языки (EN, ES, FR, DE, хинди) — Смотреть аналитику по своим роликам вместе с баерами и улучшать следующие — Темп: на старте 1–2 ролика в день, дальше рост до 5 Требования: — Портфолио законченных AI-видео, лучше рекламных или UGC — Навык сильного монтажа: ритм, динамика, хуки, удержание внимания — Владение Seedance, Kling, Higgsfield, ElevenLabs и подобными инструментами — Умение держать консистентность персонажа и связность сцен — Готовность быть на связи в субботу и воскресенье Будет плюсом: — Опыт со спай-сервисами (Tyver и др.) — Опыт с нодами / ComfyUI, автоматизации — Performance-креативы под Facebook, iGaming или арбитраж Условия: — Удалёнка, full-time — Оклад + процент от профита с креативов — Все нейросети и подписки оплачивает компания

Location: United Kingdom, London Why Sony Interactive Entertainment? Sony Interactive Entertainment isn’t just the Best Place to Play — it’s also the Best Place to Work. Sony Interactive Entertainment (SIE) is the company behind the PlayStation brand. As a subsidiary of Sony Group Corporation, we’re part of a proud legacy of innovation and excellence. SIE is a dynamic technology company, delivering cutting-edge hardware and network services to more than 100 million people and an entertainment leader, home to some of the most beloved and recognizable intellectual properties (IP) in the world. Our role at SIE is to create and nurture the experiences under the PlayStation brand, a name synonymous with entertainment excellence and creativity. About GDFT Gaming, Developer, and Future Technology (GDFT) are the technology specialists at PlayStation. We build the underlying systems that power core platform capabilities and developer tools. Our mission is to enable world-class gaming experiences by delivering low-level, high-performance technology for PlayStation platforms. The Role We are looking for an engineer to join the GDFT department, working on the design and optimisation of PlayStation’s real-time game streaming technologies. You will contribute to the development of low-latency video streaming systems across current and future PlayStation platforms, working on performance-critical areas spanning frame capture, media pipelines, codec configuration, system services and streaming transport. The role is particularly suited to engineers with a strong C/C++ background and experience working with native video or media systems, including embedded devices, TV firmware, set-top boxes, consoles, or similar. Our work sits at the intersection of product engineering and R&D. Alongside delivering production technology, engineers are encouraged to investigate new approaches to reducing end-to-end latency, improving visual quality and reducing bitrate, using measurement and experimentation to determine where meaningful improvements can be made. What You'll Be Doing Develop, optimise and maintain low-level C and C++ components within PlayStation’s game streaming stack. Investigate opportunities to reduce end-to-end streaming latency across frame capture, buffering, encoding, transport, decoding and presentation. Optimise the use and configuration of hardware video encoders, applying knowledge of codec behaviour and low-latency encoding techniques. Analyse existing media pipelines to identify unnecessary buffering, memory copies, synchronisation delays and other sources of latency. Use software and hardware instrumentation to measure system behaviour, validate hypotheses and quantify the impact of performance changes. Investigate and debug issues across user-space applications, system services and daemon-level components. Prototype and evaluate new approaches to improving streaming quality, latency and bandwidth efficiency. Work with engineers and researchers across the wider organisation on areas such as codec development, rate control, Forward Error Correction and emerging media technologies. Contribute technical findings and recommendations to design discussions, working with senior engineers to determine appropriate production solutions. Collaborate across teams to integrate and validate streaming technologies within PlayStation platform software. Attend periodic in-person workshops or team events, with travel possible a few times per year. What We Are Looking For This is a mid-level opportunity. You should bring some of the following skills and experience: Strong proficiency in C and C++. Experience developing native, performance-sensitive software rather than primarily web or managed application stacks. Experience working with real-time video, media pipelines, codecs or embedded multimedia systems. Practical understanding of video compression and media formats, including concepts such as chroma subsampling, bit depth and colour representation, with familiarity of codecs such as AVC, HEVC or AV1. Understanding of concepts relevant to low-latency video encoding, which may include GOP structures, reference frames, rate control, VBV, quantisation parameters, slices, tiles or hardware codec APIs. Strong debugging, profiling and performance-analysis skills. An ability to investigate complex technical problems, identify likely causes and use evidence and instrumentation to support conclusions. Comfortable working in unfamiliar parts of a large system and developing an understanding of how multiple components interact. An experimental mindset, with an interest in prototyping, measuring and evaluating alternative technical approaches. Ability to contribute to technical discussions and work with senior engineers when deciding how findings should be taken forward. Highly Desirable Skills: Experience in one or more of the following areas would be advantageous; we do not expect candidates to have experience across all of them: Familiarity with techniques for evaluating video codec performance, including bitrate/quality/latency trade-offs and objective or perceptual video quality metrics (such as PSNR and VMAF). Familiarity with compressed video bitstreams and codec signalling, such as NAL units/OBUs, parameter sets, profiles/levels and supplemental enhancement information. Experience developing native video players, media frameworks or codec pipelines on TVs, set-top boxes, consoles, mobile devices or other embedded platforms. Experience configuring or integrating hardware video encoders or decoders. Deeper knowledge of H.264, HEVC, AV1 or other current and emerging video codecs, beyond high-level API configuration, such as reference structures, rate-control behaviour, bitstream structure or encoder-specific optimisation. Experience optimising real-time video for low latency, visual quality or bitrate efficiency. Familiarity with frame capture, graphics presentation pipelines, framebuffers or GPU/CPU synchronisation. Experience identifying and reducing memory copies, queueing delays, memory-bandwidth limitations or other system-level performance bottlenecks. Experience working with system services, daemons, firmware interfaces or other low-level platform components. Familiarity with techniques such as content-aware encoding, region-of-interest encoding, adaptive resolution or frame rate, alternative reference-frame strategies or similar streaming optimisations. Experience with Forward Error Correction or other techniques for improving the resilience of real-time media. Experience with SIMD, vectorisation, cache optimisation or multithreaded performance tuning with instrumentation. Experience with console or embedded systems development environments. Experience with Git, Jenkins or similar CI/CD workflows. Why Work in GDFT Focused on platform-wide technology development Room to explore and innovate within performance-critical system areas Work on cross-regional collaborative projects Flexible and relaxed working environment, with strong support for hybrid work Opportunities for extensive professional development, technical conference attendance, and participation in internal hackathons Please note, Sony Interactive Entertainment conducts background checks at the offer stage for all new employees (which may include criminal background checks for some roles) and will need to process personal information to support these checks. Please refer to our Candidate Privacy Notice (where applicable) for more information about what personal information we collect, how we use it, who we share it with, and your data protection rights. Equal Opportunity Statement: Sony is an Equal Opportunity Employer. All persons will receive consideration for employment without regard to gender (including gender identity, gender expression and gender reassignment), race (including colour, nationality, ethnic or national origin), religion or belief, marital or civil partnership status, disability, age, sexual orientation, pregnancy, maternity or parental status, trade union membership or membership in any other legally protected category. We strive to create an inclusive environment, empower employees and embrace diversity. We encourage everyone to respond. Sony Interactive Entertainment is a Fair Chance employer and qualified applicants with arrest and conviction records will be considered for employment.

Location: Москва Мы — core-команда, которая отвечает за ML для понимания аудио во всем Сбере. За последний год мы выложили в open source SOTA-модель для распознавания речи на русском языке — **GigaAM**, а прошлой весной первыми в России запустили **нативное понимание звука в LLM — GigaChat Audio**. Сейчас мы активно улучшаем мультимодальные возможности GigaChat: работаем над качеством на сложных аудио- и визуальных контекстах и учим модель понимать видео не только по звуковому потоку, но и по кадрам. **Цель** — совместное обучение на vision+audio и video+audio диалогах длительностью 90+ минут. _Первый этап отбора на эту вакансию – общение с AI-рекрутером. После отклика вам на почту и в чат на платформе HeadHunter придет приглашение пройти первичное интервью с ГигаРекрутером в Telegram. Диалог займёт примерно 10 минут. Его задача — уточнить недостающие детали и ускорить рассмотрение вашей кандидатуры. ГигаРекрутер только начинает свой путь, поэтому просим относиться с пониманием. Ваш опыт и участие помогут сделать его удобным и полезным!_ Обязанности * Проводить эксперименты по обучению Vision / Audio / Video LLM: данные, стадии обучения, способы смешивания модальностей * Оптимизировать ML-инфраструктуру для обучения на длинных мультимодальных контекстах * Формулировать и проверять гипотезы о том, как улучшать качество LLM на видео. Требования * Сильный опыт в DL-инженерии, желательно в обучении LLM-моделей * Глубокое знание PyTorch, включая DDP / FSDP * Понимание distributed training и efficient deep learning: DP / TP / PP / EP / SP, mixed precision, checkpointing, offloading, профилирование и оптимизация обучения * Понимание архитектуры LLM: Transformer, attention (MHA / GQA / MLA), RoPE и другие позиционные эмбеддинги, long context, MoE * Уверенный Python на уровне production-кода: asyncio, multiprocessing, профилирование, отладка больших систем. **Будет плюсом:** * Опыт обучения мультимодальных LLM * Опыт снижения стоимости контекста при обработке видео * Опыт расширения контекста LLM * Понимание RL-подходов для обучения моделей: RLHF / RLVR, PPO / GRPO / DPO * Опыт построения бенчмарков с использованием LLM-as-a-judge. Условия * крупнейшее DS&AI community — более 600 DS-специалистов банка * дайджест о самых последних разработках в области DS&AI и отчеты с крупнейших конференций мира * возможность быть соавтором НИРов и статей для международных конференций * возможность выбрать удобный формат работы: гибрид или офис * ежегодный пересмотр зарплаты, 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 * ипотека выгоднее до 7% для каждого сотрудника * бесплатная подписка СберПрайм+, скидки на продукты компаний-партнеров * вознаграждение за рекомендацию друзей в команду Сбера.

Задачи: — создание маркетинговых креативов и playable ads для мобильных игр с использованием 2D и AI-технологий. Требования: — опыт работы с мобильными 2D и AI-креативами; — уверенное владение After Effects; — опыт работы с AI-инструментами для создания визуала и видео; — наличие портфолио с motion / video работами; — интерес к gamedev и рекламным креативам. Условия: — работа в студии CGB Studio, специализирующейся на маркетинговых креативах и playable ads для мобильных игр; — быстрый продакшен и высокая динамика работы с соблюдением строгих сроков и стандартов качества.

Location: Remote - United States Innodata (Nasdaq: INOD) is a global data engineering company. We believe that data and Artificial Intelligence (AI) are inextricably linked. Our mission is to enable the responsible advancement of artificial intelligence by providing the data, evaluation frameworks, and human expertise required to build AI systems that can be trusted at scale. We provide a range of transferable solutions, platforms, and services for Generative AI / AI builders and adopters. In every relationship, we honor our 36+ year legacy delivering the highest quality data and outstanding outcomes for our customers. Scope of the Role: Video is where multimodal models are weakest and hardest to grade. Temporal reasoning, long-form understanding, grounding events in time, and holding audio, video, and text together do not fall out of image benchmarks — and the evaluations for them are still immature. Closing that gap is gated as much by how we design data and evaluation as by architecture. Innodata builds that data and those evaluations for the customers and frontier labs advancing video and multimodal models, and we are hiring a Research Scientist to own the science behind it. You will partner directly with the customers and frontier labs building video understanding, video-language, and video-generation models, as interested in the data behind them as in the models themselves. Video spans two model families judged in completely different ways: models that understand video — answering questions, localizing events, grounding language in time — where the question is whether the answer is correct; and models that generate it, where fidelity, temporal coherence, and physical plausibility matter and no automatic metric is settled. You own the evaluation science for both, and knowing when model-based scoring can stand in for a human versus when it can't. Your conclusions shape what our partners measure and collect next. What You’ll Own: You will define how Innodata designs, structures, and evaluates video data for video and multimodal models, and you will validate those choices experimentally. Concretely, you will: Translate the requirements of video and multimodal models — video understanding, temporal and event localization, action recognition, long-form video, video-language models, video generation, cross-modal reasoning, and multimodal retrieval and grounding — into concrete data specifications: modalities, annotation schemas, sampling, and evaluation criteria. Build evaluation methodology for video understanding — temporal grounding accuracy, long-context and long-horizon reasoning, and dynamic multi-turn, cross-modal, and retrieval-and-grounding evaluation — clear about when model-based scoring is trustworthy and when a human is needed. Build evaluation methodology for video generation — fidelity, temporal coherence, and physical plausibility, including generative video used as a world model — the regime where automatic metrics are weakest and human judgment matters most. Decide how existing and incoming video should be structured, enriched, and sampled to extract the most model value from it, including from messy, domain-specific footage. Run experiments that prove data decisions matter: fine-tune and evaluate models on Innodata data, with ablations tying specific data choices to measurable improvement. Design adversarial and stumping evaluations that surface where video and multimodal systems fail, and turn those failures into better data. Publish. Turn what you learn into benchmarks, methodology, and papers that advance the field and earn the trust of the customers and frontier labs we partner with. Work with annotation teams, subject-matter experts, and the synthetic-data pipeline to turn specifications into operational collection and labeling plans. You’ll Thrive in This Role If You Have: Roughly 5+ years of hands-on industry experience in video understanding or multimodal ML. We weight practical experience over formal credentials; a PhD with a compelling, current research agenda can offset the lower end. A Bachelor's degree in computer science, electrical engineering, or a related technical or quantitative field is required; an advanced degree (MS or PhD) in a relevant field is preferred. Trained and evaluated video or multimodal models yourself, with strong PyTorch fundamentals. Fluency in the formats and tooling video work runs on: ffmpeg and decord pipelines, temporal and COCO-style annotation, WebDataset, Parquet and Arrow, and HuggingFace datasets. Experience fine-tuning large video or vision-language models with the modern toolchain (HuggingFace transformers, PEFT, efficient inference), and with long-form video, streaming, temporal segmentation, or synthetic video generation. A way of thinking in datasets and benchmarks: you have built evaluation sets, calibrated difficulty, and argued about what makes video data good for a given objective. A track record the field recognizes: first-author publications or strong open-source contributions at venues such as CVPR, ICCV, ECCV, NeurIPS, or ICLR. The ability to work directly with the research scientists at the customers and frontier labs we partner with, and to explain data and modeling decisions clearly to both expert and non-expert audiences, backed by a rigorous, reproducible approach to experiments and documentation. Bonus: interest or hands-on experience in responsible-AI evaluation and red-teaming — safety and robustness testing for video and multimodal systems. The expected salary range for this position is $160,000 - $185,000 p/year, based on experience, skills, and qualifications. Please be aware of recruitment scams involving individuals or organizations falsely claiming to represent employers. Innodata will never ask for payment, banking details, or sensitive personal information during the application process. To learn more on how to recognize job scams, please visit the Federal Trade Commission’s guide at https://consumer.ftc.gov/articles/job-scams. If you believe you’ve been targeted by a recruitment scam, please report it to Innodata at verifyjoboffer@innodata.com and consider reporting it to the FTC at ReportFraud.ftc.gov .

Location: San Francisco, CA • New York, NY • United States Figma is growing our team of passionate creatives and builders on a mission to make design accessible to all. Figma’s platform helps teams bring ideas to life—whether you're brainstorming, creating a prototype, translating designs into code, or iterating with AI. From idea to product, Figma empowers teams to streamline workflows, move faster, and work together in real time from anywhere in the world. If you're excited to shape the future of design and collaboration, join us! Video is one of the most powerful ways we bring the Figma brand to life, and we are building the function to match that ambition. As a Video Strategist, you will define what video does for this brand and build the operation that delivers it: the strategy, the franchises people come back for, the production model behind them, and eventually the team that runs it all. You will sit within Brand Marketing and partner closely with design, creative, comms, product marketing, community, and advocacy to turn complex product and customer stories into work people want to watch and share. This is a rare combination. Figma is a beloved brand with a creative community that already wants to hear from us, and a video function that is still yours to build. This is a full time role based in San Francisco or New York hub, or remotely in the United States. What you’ll do at Figma: Set the strategy: Define the role video plays across brand, social, editorial, product launch, community, and event moments. Own the format architecture and the prioritization framework that decides what we make and what we pass on Build franchises: Create and run a portfolio of recurring video programs that compound audience over time, alongside the campaign and launch work. Own the development pipeline for new formats and the call to renew, evolve, or retire a series Build the engine: Decide what we produce in-house, what goes to agencies and directors, and what we source from creators, with the economics behind each. Own the video budget and vendor relationships, and build the standards we are missing today around briefing, review, rights, and asset reuse Lead the craft: Set and hold the quality bar for video made anywhere under the Figma brand. Write and approve briefs, run creative reviews, and enable cross-functional teams producing their own video with guardrails and coaching rather than gatekeeping Measure what matters: Partner with our Brand Marketing Technologist to define what good looks like for each format and franchise, instrument it, and report it in a way that changes what we make next Grow the team: Lead, hire, and develop the video team and the external bench, and make the case for the resourcing the strategy requires We'd love to hear from you if you have: 10+ years across video, brand storytelling, editorial, content strategy, or production leadership, including several years leading a video function or team A track record creating and running recurring video franchises that outlasted a single season and grew an audience Experience owning a production model end to end: the in-house versus agency versus creator call, a real budget, and the partner relationships behind it People leadership. You have hired, developed, and given hard feedback to producers, editors, or an external bench Fluency across the format spectrum, from social-first short-form to long-form storytelling, launch films, and customer narratives, with clear opinions on what each format is good for The ability to set a creative bar for work made by people who do not report to you, and to hold it without becoming the bottleneck Comfort as the video voice in rooms with senior stakeholders, including the judgment to disagree productively and the measurement literacy to defend your decisions While it’s not required, it’s an added plus if you also have: Hands-on editing or motion graphics ability. We are not hiring you to sit on a timeline, but craft fluency sharpens judgment and helps you move fast Experience with live event and conference content at scale Familiarity with music licensing, talent releases, and usage rights at volume Experience building creator and influencer partnerships Experience scaling video inside a company going through rapid growth or a brand evolution At Figma, one of our values is Grow as you go. We believe in hiring smart, curious people who are excited to learn and develop their skills. If you’re excited about this role but your past experience doesn’t align perfectly with the points outlined in the job description, we encourage you to apply anyways. You may be just the right candidate for this or other roles. Pay Transparency Disclosure Job level and actual compensation will be decided based on factors including, but not limited to, individual qualifications objectively assessed during the interview process (including skills and prior relevant experience, potential impact, and scope of role), market demands, and specific work location. Figma offers equity to employees, as well as a competitive package of additional benefits, including health, dental, and vision coverage; retirement benefits with company contributions; parental leave and reproductive or family planning support; mental health and wellness benefits; and paid time off. Figma provides paid sick leave, holidays, and other leave benefits in compliance with applicable federal, state, and local laws, including the requirements of the Washington Minimum Wage Act and related regulations. Exempt employees are eligible for employer‑provided paid flexible PTO in addition to flexible paid sick leave. PTO is subject to manager approval. Additional benefits may include company recharge days, cell phone and home internet reimbursements, and a number of lifestyle spending accounts. Figma also offers sales incentive compensation for most sales roles and an annual bonus plan for eligible non-sales roles. All compensation and benefits are subject to applicable plan terms and may be modified by Figma at any time, consistent with applicable law. Annual Base Salary Range: $153,000 — $269,000 USD At Figma we celebrate and support our differences. We know employing a team rich in diverse thoughts, experiences, and opinions allows our employees, our product and our community to flourish. Figma is an equal opportunity workplace - we are dedicated to equal employment opportunities regardless of race, color, ancestry, religion, sex, national origin, sexual orientation, age, citizenship, marital status, disability, gender identity/expression, veteran status , or any other characteristic protected by law. We also consider qualified applicants regardless of criminal histories, consistent with legal requirements. We will work to ensure individuals with disabilities are provided reasonable accommodation to apply for a role, participate in the interview process, perform essential job functions, and receive other benefits and privileges of employment. If you require accommodation, please reach out to accommodations-ext@figma.com . These modifications enable an individual with a disability to have an equal opportunity not only to get a job, but successfully perform their job tasks to the same extent as people without disabilities. Examples of accommodations include but are not limited to: Holding interviews in an accessible location Enabling closed captioning on video conferencing Ensuring all written communication be compatible with screen readers Changing the mode or format of interviews To ensure the integrity of our hiring process and facilitate a more personal connection, we require all candidates keep their cameras on during video interviews. Additionally, if hired you will be required to attend in person onboarding. By applying for this job, the candidate acknowledges and agrees that any personal data contained in their application or supporting materials will be processed in accordance with Figma's Candidate Privacy Notice . Application limits: To help candidates focus on the opportunities most relevant to their skills and experience, Figma limits the number and frequency of applications. These limits apply to all roles: Candidates may submit up to 5 applications during any rolling 30-day period. Candidates may not reapply to the same role while an application is active or within 90 days of receiving a rejection. Once the applicable days has passed, candidates are welcome to apply again.

About the Client Our client is involved in an innovative data-collection project, which focuses on creating video datasets for training AI models. These AI models are designed to learn from first-person (POV) video footage. The captured videos depict real people involved in hands-on tasks with a mounted phone on their head or chest, paired with spoken narration. The purpose is to enable the model to connect visual inputs (objects, hands, actions) with audio descriptions. About the Role The Recorder role requires individuals who can master a specific hands-on activity from a selection of verticals. You will need to film yourself performing these tasks from a first-person perspective using your smartphone while ensuring the tasks are clearly visible. If fluent in English, you may narrate live; otherwise, ambient audio will suffice, and a separate narrator will cover the task later. Tasks & Deliverables — Select a mastered sub-area activity and record a complete end-to-end process. — Ensure hands and the procedure are constantly in focus. — Provide live narration if fluent in English or ambient audio for post-recording narration. — Deliver videos in MP4 format with accompanying JSON metadata. — Maintain narration density at 25 words per minute; ensure 90%+ of speech describes the setting or task steps. Required Skills — Mastery of at least one specific sub-area activity from the provided list (e.g., cleaning, laundry, organizing, pet care, repairs, gardening, sewing, woodworking, automotive repair, art, technology use, personal care, medical procedures, culinary skills, lab protocols). — Capability to record stable, clear POV videos with a smartphone. — Ability to work in a real setting without third parties in frame. — Competent in following a prepare → execute → finish script for filming tasks. Nice-to-have Skills — Experience recording tutorials or classes. — Access to various settings for diverse video backgrounds. — Owns equipment like a head/chest mount, image stabilization tools, or a good microphone. Rules — No personal data on video — no third-party faces, screens, documents, plates, addresses, mirror reflections. — 100% human narration — no TTS or synthetic voice. Engagement Details — Compensation model: Hourly paid, considering the total number of hours for approved videos. — Commitment Type: Flexible hours based on video submissions. — Duration: Ongoing, with earnings based on the number of approved videos. — Location: Remote, with flexible overlap with client's timezone. — Start Date: Immediate start upon approval of test video.

About the Client: Our client is involved in an innovative data-collection project, which focuses on creating video datasets for training AI models. These AI models are designed to learn from first-person (POV) video footage. The captured videos depict real people involved in hands-on tasks with a mounted phone on their head or chest, paired with spoken narration. The purpose is to enable the model to connect visual inputs (objects, hands, actions) with audio descriptions. Sub-areas for the videos include: Home & Daily Tasks: cleaning, laundry (sorting, folding), organizing a closet, house tours, pet care, packing luggage, loading appliances / dishwashing. Repairs & DIY: home repair, gardening / farming, furniture assembly, woodworking, plumbing, electrical work, bicycle maintenance, soldering electronics. Textile & Craft Arts: sewing, knitting, crocheting, using a loom, leathercrafting, bookbinding, crafting jewelry, pottery / ceramics. Professional Trades & Industrial: automotive repair / maintenance, warehousing / logistics, construction / woodworking, laboratory work, operating heavy machinery controls, assembly-line packaging. Hobbies & Arts: art – drawing, ceramics, painting, playing instruments, model building, outdoor survival / camping, calligraphy. Technology & Computing: using computers or devices (e.g. audio mixers), gaming, product demos, VR / AR interaction, 3D printer setup / maintenance. Outdoors & Activity: city tours, shopping, navigating public transit. Personal Care: haircut, applying makeup, detailed grooming routines. Specialized & Professional: medical procedures, first aid training, professional barista workflows, culinary chef / knife work, lab protocols (pipetting, titrations). About the Role - Narrator: Watches the POV video and describes the scene and task in first person, in English, as if explaining it to a model that can't see. Fluency and describing ability matter most; knowing the activity is not required. Deliverable: — audio narration synced to the video, recorded after the footage. Language: — English; fluency is criterion #1. Density: — at least 25 words/min on average, no long silences. Content: — 90%+ about the setting or task steps. Required Skills: — English fluency. — Ability to narrate pre-recorded videos. — All details on the Acceptance Criteria will be shared prior to the work. Rules: — No personal data on video — third-party faces, screens, documents, plates, addresses, mirror reflections. — 100% human narration — no TTS or synthetic voice. Engagement Details: Compensation model: Hourly paid, considering the total number of hours for approved videos. Commitment Type: Flexible hours based on video submissions. Duration: Ongoing, with earnings based on the number of approved videos. Location: Remote, with flexible overlap with client's timezone. Start Date: Immediate start upon approval of test video.