Join the mission to make communications trustworthy in the age of AI
About Us:
is a Zurich-born startup on a mission to restore trust in digital communication by protecting high stakes organizations from AI-generated voice fraud in real-time. Banks, Government agencies, and enterprises use our deepfake-detection to stop account take-overs, CEO impersonations and fake news before they cause damage.
We're a team of passionate innovators, coming from top tier tech and consulting firms as well as startups. We thrive on collaboration, embrace a "move fast to launch and iterate" mentality, and are united by a shared vision of a safer, more transparent world in the age of AI.
Why we need you:
We are pioneering cryptographic audio watermarking: a future-proof solution designed to break free from the cat-and-mouse game of deepfake detection. Instead of chasing artifacts in synthetic voices, our approach embeds authenticity directly into the audio signal at the point of capture, ensuring trust regardless of format or compression. As a research scientist on our team, you will help turn this vision into reality: designing the core algorithms, stress-testing them against adversarial attacks, and shaping the models that will become the backbone of trusted voice communication worldwide.
What you will do:
* Build the foundation:
Train, benchmark and modify the architecture of machine learning models, for robust audio watermark embedding and detection under real-world conditions (compression, re-encoding, noise, streaming).
- Stress-test robustness:
Design adversarial scenarios and run large-scale experiments to validate watermark resilience against attacks.
- Bridge research and product:
Translate academic insights into production-ready code, and watermarking pipelines that can scale.
- Push the frontier:
Explore novel approaches at the intersection of signal processing, machine learning, and security.
About you:
* Experience:
2–5 years in data science or machine learning, ideally with a research background.
- Technical depth in ML & audio:
Proficient in Python, PyTorch, and modern ML frameworks, with experience in large-scale experiments and a solid grasp of audio signal processing (or eagerness to learn).
- Research mindset:
Curious, methodical, and able to critically evaluate ideas from both academia and industry.
- Ownership & team spirit:
You're self-driven, love solving complex problems, can run projects end-to-end, and contribute to a positive and fun team environment.
- Languages:
Fluent in English. Any other language is a plus.
Nice-to-have:
* Prior experience in audio watermarking, steganography, or multimedia security.
* Publications or contributions in machine learning and signal processing.
* Familiarity with adversarial ML and robustness testing.
* Experience with real-time or streaming audio systems.
* Contributions to open-source ML or audio research projects.
Benefits:
* Make a real impact:
Join a company that is working with cutting-edge machine learning and fighting the good fight.
- Be part of something big:
Help protect the world from AI-generated chaos and shape our product and culture from the beginning.
- Work with cool people:
Our team is smart, passionate, and driven.
- Flexibility: Work from wherever you're most productive and collaborative. (At least two days a week in the office in Zurich). Part-time possible (80%-100%).
- Competitive salary and sharing the value you help create:
We're offering a meaningful ownership stake in the company. CHF ~80k salary + ~CHF 15k in stock options (depending on role and experience).
If you're ready to use your skills for good and want to be part of a team from the start that's making a real difference, we want to hear from you