Share these talks and lectures with your colleagues
Invite colleaguesCase study
Empowering content search: Leveraging the potential of machine learning
Journal of Digital Media Management,
12
(2),
116-126
(2023)
Abstract
In the digital age, the power of machine learning is harnessed to transform content search capabilities. This integration heralds a new era for content creation tools, granting editors and creators access to granular, frame-level content insights. These capabilities enable precision adjustments, enhancing the final product’s efficiency and quality. Recent years have witnessed remarkable shifts in using machine-generated data within media tools. However, harnessing the full potential of machine-learning techniques poses challenges due to the vast and diverse data generated by many algorithms. In response, Netflix has pioneered a groundbreaking solution: the Media Understanding Platform. This platform is a unifying abstraction layer across all Netflix studio applications, bridging the gap between client and machine-learning platforms. This paper illustrates the platform’s design and prowess through real-world examples of promotional media tools that enrich content discovery within Netflix’s expansive catalogue, offering a glimpse into the future of content search.
Keywords: digital assets search; machine learning; creative tools; standard schema; centralised platform
The full article is available to subscribers to the journal.
Author's Biography
Meenakshi Jindal is a staff software engineer specialising in content infrastructure and solutions at Netflix Inc. Since 2018, she has been involved in shaping the architecture and execution of the company’s media asset management platform. Her expertise is in designing digital asset management components utilising distributed system architecture for content production and post-production workflows. She has actively collaborated with studio partner applications to ensure the platform’s evolution aligns with the dynamic shifts in the media industry.
Varun Sekhri is a staff software engineer at Netflix Inc. As part of the Asset Management Platform team, he works on many projects, including digital asset management, annotation service and workflow management of assets. Before Netflix, he worked at Microsoft and Uber. He holds a master’s degree from Iowa State University. His interests include the architecture of large-scale systems and distributed computing.
Tiffany Low is a senior software engineer within the creative innovation team at Netflix Inc. The team develops specific and actionable insights to improve the promotional strategies and asset suites for the Netflix member product, content strategy and member base. Together with the other members of the initiative, she has been working on onboarding more insights and more product workflows to the media understanding platform.
