Applied Scientist at Flawless

You will bridge the gap between research and production by scaling audio/video dataset generation and refining end-to-end multimodal pipelines. This role is critical for operationalizing lip-sync work, owning model training workflows, and developing rigorous evaluation metrics to ensure high-fidelity releases that meet the demanding quality standards of the film industry.

• Develop and scale repeatable audio/video dataset curation pipelines and lip-sync model training workflows across diverse datasets.
• Design and automate quantitative and qualitative metric testing pipelines to evaluate and visualize model performance and audio-visual quality.
• Partner with researchers and VFX artists to validate model improvements and align technical outputs with ambitious company product goals.

• Holds an MSc or PhD in Audio Processing, 3D Computer Vision, Speech Synthesis, or a related multimodal field with industry experience.
• Demonstrates expertise in Python and deep learning frameworks like PyTorch, with a strong foundation in math and signal processing.
• Possesses a product-driven mindset and experience developing large-scale data processing systems for generative or audio-visual speech applications.

• Work at the cutting edge of generative AI for film, building systems that directly influence the future of global storytelling and visual effects.
• Join a high-caliber team where your work translates technical novelty into tangible product value, improving out-of-the-box approval rates for creative professionals.
• Gain unique exposure to multimodal fusion and neural speech synthesis, operating in a high-impact environment that combines academic rigor with rapid production cycles.