Job Description
Research Scientist (Deepfake)
Remote US and Canada
Who we are
As human beings, one of our fundamental identifiers is our voice. Pindrop’s advanced voice identity technology recognizes this distinct and unique human quality with the kind of precision and certainty that’s needed when information or access is essential. From preventing fraud in call centers to obtaining information from smart devices and even activating cars, Pindrop lets people use their voice to quickly and privately connect to, enter, and unlock their world. At Pindrop, we hire great people and we take care of them. All we do is guided by our Core Values: Audaciously Innovate, Evangelical Customers for Life, Execution Excellence, Win as a Company, Make a Difference.
Headquartered in Atlanta, GA, Pindrop has raised over $223M in capital by premier VCs including Andreessen-Horowitz, IVP, and CapitalG.
We are looking for a Research Scientist/Engineer to join the Speech team to help us explore exciting research ideas in the area of audio-visual deepfake detection. This candidate will work in a world-class team of researchers and will play a key role in developing the next generation of voice and video experiences.
What you’ll do
- Conduct experimental studies on various audio and computer vision tasks to assess accuracy and quality of pre-existing models
- Help the transition of research models into production on different platforms including embedded systems
- Conduct proof-of-concept experiments on lab and customer data
- Publish patents and peer-reviewed papers in top audio and computer vision conferences
- Help your teammates in reviewing their research projects
- Propose new ML models that improve Pindrop products
Who you are
- You are a creative problem solver who is excited about building amazing products that add real value to millions of people
- You are enthusiastic about voice security as well as developing research ideas into products
- You can explain complex topics in simple terms, and you love building strong relationships with colleagues and stakeholders
- You are a self-starter and excel in a fast-paced dynamic environment that often includes ambiguity
Your skill-set:
- Master and/or PhDs in a quantitative field (Computer Science, Mathematics, Engineering, Artificial Intelligence, etc.)
- 1+ years of professional experience as a researcher in voice security, audio/video deepfake detection, speech synthesis, speaker recognition and/or generative AI.
- Experience with audio and/or video deepfake detection required
- Strong Python skills
- Experience with ML frameworks: PyTorch, TensorFlow, Keras
- Proven track record of peer-reviewed publications at prestigious conferences and journals.
- Proven track record of successful and timely project delivery
- Strong collaboration & proven success in partnering with stakeholders
- Preferably, you have experience with ML/Speech challenges (ASvspoof, Kaggle, NIST SRE, Chime, etc.)
- Preferably, you have experience with C/C++
- Preferably, you have deep knowledge of biometrics, authentication, fraud, or customer experience concepts
- Preferably, you have experience communicating with internal and external customers around proof-of-concepts, issues, etc.