About the Role

Synthesia has already revolutionized how tens of thousands of businesses create video, and now we are setting our sights on something much larger. We are designing the next generation of human interactive AI, creating systems that perceive, comprehend, and respond naturally across video, audio, and language simultaneously. As a Staff Research Engineer in Multimodal Generative Modelling, you will take the lead on solving foundational deep learning problems that bridge the gap between static generation and fluid, real-time multimodal interaction. This is a fully remote, worldwide opportunity that gives you the freedom to conduct cutting-edge research from anywhere on the globe.

Overview

Following our recent 200 million dollar Series E funding round, Synthesia reached a valuation of 4 billion dollars, supported by top tier investors such as NVentures, Kleiner Perkins, Accel, and GV. More than ninety percent of the Fortune 100 relies on our software to communicate visual ideas. To power the next era of enterprise communication, we are building unified generative models capable of processing and producing rich audiovisual output on the fly. You will collaborate directly with world-class scientists and engineers, driving the core machine learning innovations that power interactive digital humans.

Key Responsibilities

In this role, you will shape the future of interactive AI by taking on several core focus areas

Architect, train, and scale novel generative models that weave together video, speech, and textual signals into coherent real-time experiences
Tackle fundamental challenges around latency, multimodal alignment, dynamic expression generation, and generation fidelity
Design robust training and evaluation pipelines across massive multimodal datasets using modern distributed compute clusters
Work closely with applied research teams and product engineers to bring theoretical breakthroughs directly into production applications
Publish research, file patents, and establish Synthesia as the undeniable leader in interactive generative media

Requirements

We are searching for accomplished researchers and practitioners who bring deep technical knowledge and a relentless drive to explore the unknown

Extensive background in machine learning research, specifically within generative modeling, computer vision, speech synthesis, or multimodal foundation models
Proven track record of designing and training large scale neural networks using PyTorch, JAX, or similar frameworks
Demonstrated experience dealing with massive compute environments, distributed training paradigms, and performance optimization
Solid publication record at top tier venues such as CVPR, ICCV, NeurIPS, ICML, or ICLR, or equivalent high-impact industry contributions
Strong communication skills and the self-motivation required to drive complex research initiatives in an autonomous, global team

What You Will Gain

Joining Synthesia at this pivotal stage offers exceptional benefits and growth opportunities

A genuinely remote-first work culture that lets you work from virtually anywhere in the world
Uncapped access to substantial compute resources and cutting-edge hardware to run ambitious experiments
Direct impact on a product used daily by millions of viewers and over sixty thousand businesses worldwide
A highly competitive compensation package including meaningful equity in a profitable, rapidly expanding company
An intellectually rich environment surrounded by passionate peers dedicated to humane, useful artificial intelligence