Reality Defender

Applied Scientist Intern (Audio Language Modeling)

New York, New York, United States

Reality Defender Logo
Not SpecifiedCompensation
InternshipExperience Level
Full TimeJob Type
UnknownVisa
AI & Machine Learning, CybersecurityIndustries

Requirements

Candidates must be currently enrolled in a PhD program with a specialization in machine learning, deep learning, natural language processing, or speech processing. They should have at least 2 years of experience in training and fine-tuning large models, particularly audio language models and multi-modal foundation models. Familiarity with the end-to-end model building pipeline for machine learning tasks, including dataset curation, model implementation, benchmarking, and result analysis, is required. Experience with distributed multi-GPU training and prior publications in reputable ML, audio, or NLP research venues is also necessary.

Responsibilities

The Applied Scientist Intern will explore and conceptualize novel methods to leverage different modalities for deepfake detection and audio understanding tasks. They will perform fundamental and applied research to advance the state-of-the-art in audio deepfake detection and build models that generalize to unseen generative methods. Collaboration with scientists and engineers across the organization is expected, as well as summarizing, publishing, and presenting research findings.

Skills

Machine Learning
Deep Learning
Natural Language Processing
Speech Processing
Audio Language Models
Speech Foundation Models
Multi-modal Foundation Models
Distributed Multi-GPU Training
Model Building
Dataset Curation
Benchmarking
Result Analysis

Reality Defender

Deepfake detection for enterprises and governments

About Reality Defender

Reality Defender offers deepfake detection solutions to protect enterprises, platforms, and governments from AI-generated threats. Its detection platform scans images, videos, and audio in real time to identify fabricated content, helping to prevent misinformation. The company stands out by providing enterprise-grade services through a subscription model that allows easy integration into existing systems. The goal is to enhance fraud prevention and maintain the authenticity of digital content for clients.

Key Metrics

New York City, New YorkHeadquarters
2018Year Founded
$46.7MTotal Funding
SERIES_ACompany Stage
Enterprise Software, Cybersecurity, AI & Machine LearningIndustries
51-200Employees

Risks

Free tools like TrueMedia's may undercut Reality Defender's subscription model.
Rapid increase in deepfakes could overwhelm current detection capabilities.
Commoditization of AI tools challenges Reality Defender to update detection algorithms.

Differentiation

Reality Defender offers real-time detection for images, videos, and audio deepfakes.
The platform is government-approved, ensuring high reliability and accuracy for clients.
Reality Defender uses a multi-model approach, enhancing detection capabilities across various media types.

Upsides

Raised $33M in Series A to expand technology and market reach.
Partnership with Respeecher enhances audio deepfake detection capabilities.
Won 'Most Innovative Startup' at RSA Conference 2024, boosting credibility.

Land your dream remote job 3x faster with AI