Jack Saunders, PhD

Co-Founder

Chippenham, England, United Kingdom1 yr 8 mos experience

Most Likely To SwitchAI Enabled

Key Highlights

PhD in Artificial Intelligence with extensive research experience
Founder of RealSync AI, specializing in Avatar technology
Published research in top venues like CVPR and Eurographics

Stackforce AI infers this person is a specialist in AI and digital human technologies within the SaaS industry.

Contact

Skills

Core Skills

Gaussian SplattingDiffusion ModelsGenerative AiDigital AvatarsArtificial Intelligence (ai)

Other Skills

VLMLarge Language Models (LLM)Multimodal AIConversational AIvideo diffusionMachine LearningNeural RenderingReport WritingTechnical WritingWritingPython (Programming Language)TransformersGrant WritingGenerative Adversarial Networks (GANs)Deep Learning

About

I'm a Research Scientist, working on digital humans, with a PhD from the University of Bath and Microsoft internship experience. My research intersects artificial intelligence, mixed reality, vision and graphics, and has been published in top venues, (CVPR, Eurographics, BMVC). I'm also the founder of RealSync AI, where I perform consultanacy services in the Avatar space. Through my work in the Avatar space, I have hands-on experience with Generative AI (including GANs, Diffusion Models and Flow Matching), transformers, neural rendering (Neural Textures, NeRFs, Gaussian Splatting), inverse rendering and optimization. I have extensive experience in frameworks such as Python, Pytorch and Pytorch3D.

Experience

Tavus

2 roles

Senior Research Scientist

Promoted

Jan 2026 – Present · 3 mos · United Kingdom · Remote

Leading the exploration of new types of models for video generation including Gaussian Splatting based models, flow matching, diffusion, distillation techniques and much more. I am also assisting on projects relating to conversational modelling with VLM/LLMs, visual perception, turn taking models and everything else involved in making Avatars feel natural and conversations realistic.

Research Scientist

Feb 2025 – Jan 2026 · 11 mos · United Kingdom · Remote

I am helping Tavus develop the technology behind the human layer of AI:
Worked on the Phoenix class of models through two releases (Phoenix-3 and 4).
Developed an ultra-fast Gaussian Splatting model with a greater level of micro and marco expressions
Used temporal, autoregressive diffusion models to achieve an emotionally expressive Avatar with a 16% increase in lip sync measure by AVSR.

Gaussian SplattingVLMLarge Language Models (LLM)Diffusion ModelsMultimodal AIConversational AI+1

Realsync ai

Founder

Jun 2024 – Present · 1 yr 10 mos · Bath, England, United Kingdom · Remote

I founded RealSyncAI, a consulting firm specialising in AI lip-sync technology. By offering services in Avatar/Dubbing/Lip-Sync Models, I help companies stay competitive in the market.
I have consulted with more than 7 companies, ranging from pre-seed to organisations with over 100 people.
I offer strategic reports, technical implementation details, advising services for founders and engineers, and help with product market fit.
I also publish regular content in the form of social media posts, blogs and newletters, covering the digital human space and offering insights.

Generative AIMachine LearningArtificial Intelligence (AI)Digital AvatarsNeural RenderingReport Writing+2

Microsoft

Research Scientist Intern

Jun 2024 – Sep 2024 · 3 mos · Seattle, Washington, United States · On-site

As a Research Intern at Microsoft in Seattle, I had the opportunity to work on cutting-edge technology in the Mixed Reality and Cloud organisation.
I developed a Gaussian Splatting based Avatar model that needed only a single image of a user, trained on purely synthetic data for ease of use and privacy.
My work was published in CVPR2025 and developed into a working prototype that ran in realtime on a Surface Laptop.
I also contributed ideas to another interns work using LLMs for spatial reasoning to appear in AAAI26.

Generative AIMachine LearningArtificial Intelligence (AI)Digital AvatarsNeural Rendering

Deepreel

AI Research Scientist

May 2022 – May 2024 · 2 yrs · London Area, United Kingdom · Remote

In my role as an AI Scientist at DeepReel, I spearheaded the creation of a cutting-edge text-to-avatar system utilizing advanced technologies like 3D Morphable Models, Neural Rendering and Transformers. I also developed prototypes utilizing a wide range of technologies such as NeRFs/Gaussian Splatting and Diffusion Models. Additionally, I successfully secured grant funding for a project aimed at enhancing data efficiency in AI Avatars, showcasing my ability to drive innovation and secure resources for impactful projects.

Generative AIMachine LearningPython (Programming Language)Artificial Intelligence (AI)TransformersDigital Avatars+3

Epic games

Research Scientist Intern

Nov 2021 – Jun 2022 · 7 mos · Manchester, England, United Kingdom

I proposed and developed a system for animating digital humans by altering their personal style and emotional characteristics with RNNs and GANs. This involved collaborating with cross-functional teams to implement cutting-edge technology for realistic digital human animations.

Generative AIMachine LearningArtificial Intelligence (AI)Digital AvatarsGenerative Adversarial Networks (GANs)Deep Learning+1