Applied Research Scientist, AI Research
Descript · San Francisco; Hybrid or Remote
About this role
**Applied Research Scientist, AI Research** **About the Role** Descript's Research team builds generative video and audio models, and multimodal understanding systems behind features like Video Regenerate, lipsync, video translation, and zero-shot voice cloning. We're a small, senior-heavy team focused on research that ships—most work goes from prototype to production features used by millions within months. **What You'll Do** • **Generative media synthesis** — Build models for photorealistic video and audio synthesis (Video Regenerate, lipsync, video translation) • **Zero-shot voice and roomtone cloning** — Realistic cloning from minimal reference audio • **Multimodal understanding** — Vision-language systems for agentic editing features • **Computer vision** — Digital human reconstruction, facial modeling, and related work • **New algorithms** — Media synthesis, speech enhancement, anomaly detection, audio/video tagging • **Direction-setting** — Identify and pursue research directions that become Descript features **What You Bring** **Required:** • Proven ability to design and implement deep learning algorithms (publications, open-source, shipped models) • Strong programming skills and deep fluency in PyTorch and/or TensorFlow • Track record of generating new ML ideas—you produce more ideas than you can implement • Strong experimental judgment and honest evaluation of results • PhD/Master's in deep learning or equivalent experience **At least one of:** • Lead/first author publication at top venue (NeurIPS, ICML, ICLR, ICASSP, ICCV, CVPR, Interspeech, SIGGRAPH, etc.) • Key role shipping a production feature with deep learning as core component **Valuable expertise:** • Generative modeling for video, audio, or images • Vision-language models and multimodal understanding • Speech and audio modeling (compression, enhancement, synthesis) • Evaluation systems for generative/agentic outputs • Taking research ideas to shipped production features **Compensation** Base salary: $261,625–$299,000, plus equity and benefits. Final offer considers experience, expertise, location, and level. **About Descript** Descript is building an AI-powered editing tool for video and audio. 150-person team backed by OpenAI, Andreessen Horowitz, Redpoint Ventures, and Spark Capital. Headquarters in San Francisco's Mission District with remote and hybrid options. Benefits include healthcare, 401k matching, catered lunches, and flexible vacation. Equal opportunity employer.
Listing freshness
CronJobs last confirmed this listing 1h ago. If its source stops confirming the opening for seven days, this page is removed from active inventory.