Staff AI Engineer (Audio)

Role Overview

Join Reality Defender as a Staff AI Engineer (Audio). You will build, tune, and deploy audio deepfake detection models, working cross-functionally with clients and internal teams to ensure robustness, reliability, and high performance in production.

Key Missions & Responsibilities

• Design, build, and optimize production-scale ML/DL models for audio deepfake detection, ensuring robustness across diverse real-world conditions.
• Partner with clients to understand production environments, define model performance criteria, investigate failure cases, and implement mitigation strategies.
• Monitor and measure model performance in production using data analytics and AI observability tooling; detect degradation trends and emerging threat patterns.
• Build and maintain dashboards to report performance and drive improvements.
• Collaborate with cross-functional partners to deploy scalable, production-grade models.

Profile / Requirements

• Deep hands-on expertise in model building, training, benchmarking, and productionization.
• Ability to translate complex model behavior into actionable insights for both technical and non-technical stakeholders.
• PhD in Computer Science, Machine Learning, Signal Processing, or a related field (or equivalent industry experience).
• Strong understanding of audio processing, classification/detection metrics (e.g., ROC, DET curves, precision/recall trade-offs), and model robustness/evaluation.
• Proven experience with large-scale training and benchmarking pipelines.
• Client-facing or solutions-oriented AI experience is a plus.
• Strong Python programming skills and ML frameworks such as PyTorch and JAX.
• Preference for candidates comfortable with EST work hours.
• 2+ years of industry experience building and deploying ML/DL models for AI products.

Nice-to-haves

• Client-facing solutions experience.
• Comfort working with EST time zone schedule.