Computer Vision Engineer

📋 Description

• • Build and deploy multimodal AI detection systems that identify AI-generated images using deep learning models, owning the full lifecycle from data pipelines and training to production deployment, monitoring, and retraining.
• • Design and operate scalable production infrastructure for model inference services, batch pipelines, and serving systems with strict attention to latency, throughput, cost efficiency, and reliability.
• • Develop evaluation and monitoring pipelines to detect model drift, performance regressions, and emerging adversarial techniques as new generative models evolve.
• • Create and manage synthetic dataset generation systems and large-scale data processing pipelines to support model training and validation.
• • Improve model accuracy, robustness, and computational efficiency through iterative experimentation grounded in real-world production metrics and user feedback.
• • Collaborate with engineering and product teams to translate trained models into shipped product features, ensuring seamless integration and user impact.
• • Communicate technical results and research findings to both technical and non-technical audiences through documentation, whitepapers, blog posts, and internal presentations.
• • Maintain and enhance the company’s AI detection capabilities against a rapidly changing threat landscape of generative AI tools and techniques.
• • Work in person in Brooklyn, New York, as part of a small, high-collaboration team that prioritizes in-person interaction for innovation and technical decision-making.
• • Contribute to the company’s research output by publishing state-of-the-art detection techniques and advancing the field of AI authenticity verification.
• • Ensure all deployed systems are continuously improved through rigorous testing, feedback loops, and operational discipline.
• • Own the technical direction and strategic execution of core detection systems with high autonomy and accountability.
• • Participate in cross-functional initiatives to align model performance with product goals and customer needs.
• • Stay current with advancements in generative models, including diffusion models, GANs, and autoregressive image architectures, and apply insights to detection system improvements.
• • Implement and optimize neural network architectures for image, video, audio, time series, and signal data to enhance detection fidelity across modalities.
• • Leverage PyTorch for large-scale distributed training of deep learning models in production environments.
• • Integrate model observability, CI/CD practices, and MLOps tooling to ensure reliable and repeatable deployment cycles.
• • Contribute to the development of adversarial robustness techniques and interpretability methods to increase trust and transparency in detection outcomes.