Applied Research Scientist, AI Research

Descript
San Francisco
Workplace: HybridFull timeUSD 216,125 - 262,438Function: Data Science & Machine LearningEducation: phdSkills: ["Experimental judgment","Clear communication","Idea generation","Evaluation-driven experimentation"]

Build and ship generative media and multimodal AI systems for video and audio editing. You’ll develop models powering features like Video Regenerate, lipsync, and video translation, plus zero-shot voice and roomtone cloning. Work on computer-vision-heavy problems such as digital human reconstruction and facial modeling, and drive new research directions from experiments to production-ready features with strong evaluation and judgment.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
Descript
Descript
5 days ago

Applied Research Scientist, AI Research

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 8 hours agoStatus: Live

Job Summary

Build and ship generative media and multimodal AI systems for video and audio editing. You’ll develop models powering features like Video Regenerate, lipsync, and video translation, plus zero-shot voice and roomtone cloning. Work on computer-vision-heavy problems such as digital human reconstruction and facial modeling, and drive new research directions from experiments to production-ready features with strong evaluation and judgment.
Location: San Francisco
Workplace: Hybrid
Employment Type: Full time
Job Function: Data Science & Machine Learning

Key Responsibilities

  • •Build generative media synthesis models that power production features like Video Regenerate, lipsync, and video translation.
  • •Develop systems for zero-shot voice and roomtone cloning from only a few minutes of reference audio.
  • •Create multimodal understanding systems and evaluation pipelines to balance quality against cost and latency.
  • •Solve computer-vision-heavy problems such as digital human reconstruction and facial modeling for more natural editing tools.
  • •Own new research directions and take ideas through framing, experiments, analysis, and production-facing feature outcomes.

Pay and Benefits

Salary: USD 216,125 - 262,438
Equity and Bonus:Equity
Perks:Health Insurance401kMeal AllowancePaid Leave

Key Requirements

  • •Proven ability to design and implement deep learning algorithms, shown via publications, open-source work, or models shipped.
  • •Strong programming skills and deep fluency in PyTorch and/or TensorFlow.
  • •Track record of generating new ML ideas and running/evaluating many experiments quickly once setups are in place.
  • •Strong experimental judgment and honesty about what does and doesn’t work.
  • •PhD or Master’s in deep learning or a related field, or equivalent experience; plus evidence as lead/first author in top venues or a key role shipping production deep-learning features.
Experience:Deep learningGenerative modelingComputer visionMultimodal AISpeech/audioProduction ML
Education:PhD / Doctorate in deep learning or a related field
Skills:Experimental judgmentClear communicationIdea generationEvaluation-driven experimentation
Languages:English
Tech Stack:PyTorchTensorFlow

Company Brief

Descript
Descript provides an all-in-one audio and video editing platform that combines transcription, multitrack editing, screen recording, and AI-powered tools for creators and teams to produce podcasts, videos, and marketing content more quickly and collaboratively.
Industry: Enterprise Software
Company Size: Large (251 to 1,000 employees)
Growth: Scaleup
Funding: Series C
Headquarters: San Francisco, United States
Founded: 2017
WebsiteLinkedIn