Full-Time - AI Application Engineer

LocationTokyo - Osaki

SenioritySenior

Company size10,000+ people

PostedDec 29, 2025 · 9mo ago

Verified live12h ago

At a glanceSummarised by Seekless from the posting.

Master's degree in Computer Science or a related field

Strong research or engineering background in machine learning, computer vision, or generative models

Hands-on experience with diffusion models, GANs, transformers, or related deep learning architectures

Proficiency in Python and deep learning frameworks such as PyTorch or TensorFlow

Demonstrated publication record in top-tier conferences or journals (e.g., CVPR, ICCV, NeurIPS, ICLR, ICML, SIGGRAPH)

Ability to design, train, and evaluate large-scale models or pipelines

Strong problem-solving skills and a passion for tackling open research challenges

Ability to communicate effectively in English (both written and spoken)

Ph.D. in Computer Science or a related field

Experience in image or video generation for creative applications (e.g., anime, film, gaming, VFX)

Experience with dataset design, annotation, and curation for generative AI

Understanding of creative production workflows (animation pipelines, film editing, compositing, or game asset creation)

Hands-on experience with large-scale training on distributed systems or cloud platforms

Strong track record of collaboration with interdisciplinary teams (e.g., artists, designers, producers)

Interest in Japanese culture, anime, or entertainment industries

Japanese language skills are a plus, but not required

Skills

Python

PyTorch

TensorFlow

JavaScript

AWS

GCP

Azure

Slack

GitHub

MS Teams

Technology Field

Machine Learning

Requirements

Computer Vision

Position Summary

At Sony R&D, we are pioneering the next generation of generative AI technologies that will transform how creators and audiences experience entertainment. From anime to film, music to games, we are working across the Sony Group to build AI that doesn’t just push the boundaries of research, but also empowers creators to tell their stories in new and powerful ways.

Here are two of the groundbreaking projects you could be part of:

■ Video Generation AI

We are also venturing into one of the most exciting frontiers of generative AI: video generation. Our goal is to develop technologies that can revolutionize how movies, music videos, games, and anime are produced across Sony’s global businesses.

Current video generation models show great promise but are far from production-ready. To make them truly usable, we need to solve critical challenges such as:

•

Controllability: camera work, character motion, lip-sync, preventing subject fragmentation

•

Editability: video inpainting, style transfer, compositing

•

Customization: tuning for specific IPs, ensuring character consistency, adjusting motion dynamics, reinforcing models with creator feedback

This project is about building video AI that creators can actually use to tell compelling stories—not just generating flashy demos.

If you are excited about building the future of video creation and transforming the global entertainment industry, we’d love to have you on board.

■ Anime-Focused Image Generation AI

Japan is home to one of the world’s richest collections of anime IP, including those from Sony’s group companies. We are taking on the bold challenge of developing a next-generation foundation model specialized in anime image generation—an AI that can truly transform how anime is made.

Existing reference-based editing AIs may deliver strong draftsmanship, but they often fail to preserve artistic style and IP consistency, making them unsuitable for real production. Our vision is to go beyond what looks technically impressive and instead create an AI that creators genuinely need and can rely on in practice.

This project is not only about pushing diffusion models—it’s about crafting high-quality datasets, extending existing models, and designing the full production pipeline required to make AI actually usable in animation studios.

If you’ve ever dreamed of shaping the future of anime with AI, this is your chance to lead the way.

Responsibilities

■ Conduct research and development on state-of-the-art image and video generation AI models

■ Design and implement datasets, pipelines, and tools tailored for creative production workflows

■ Collaborate with creators, studios, and partner companies to translate research into practical solutions

■ Publish findings at top-tier international conferences and contribute to Sony’s global R&D leadership

■ Drive innovation that enables new experiences across anime, film, music, gaming, and beyond

Required Qualifications

■ Master’s degree in Computer Science or a related field

■ Strong research or engineering background in machine learning, computer vision, or generative models

■ Hands-on experience with diffusion models, GANs, transformers, or related deep learning architectures

■ Proficiency in Python and deep learning frameworks such as PyTorch or TensorFlow

■ Demonstrated publication record in top-tier conferences or journals (e.g., CVPR, ICCV, NeurIPS, ICLR, ICML, SIGGRAPH)

■ Ability to design, train, and evaluate large-scale models or pipelines

■ Strong problem-solving skills and a passion for tackling open research challenges

■ Ability to communicate effectively in English (both written and spoken)

Preferred Qualifications

■ Ph.D. in Computer Science or a related field

■ Experience in image or video generation for creative applications (e.g., anime, film, gaming, VFX)

■ Experience with dataset design, annotation, and curation for generative AI

■ Understanding of creative production workflows (animation pipelines, film editing, compositing, or game asset creation)

■ Hands-on experience with large-scale training on distributed systems or cloud platforms

■ Strong track record of collaboration with interdisciplinary teams (e.g., artists, designers, producers)

■ Interest in Japanese culture, anime, or entertainment industries

■ Japanese language skills are a plus, but not required