The new Pinterest AWS deal represents a 4 billion dollar commitment to cloud computing services through 2031. This agreement marks the largest infrastructure investment in the history of the visual discovery platform. Consequently, the company aims to accelerate its artificial intelligence capabilities for over 600 million monthly active users.
The Pinterest AWS deal expands a long-standing relationship between the two companies that began in 2010. Over the past decade, they worked together to optimize data storage systems. Meanwhile, the renewed collaboration will support the next phase of growth in model training, inference, and platform infrastructure.
Details of the Pinterest AWS deal
Under the terms of the agreement, the visual search platform will use specialized cloud resources to train and deploy machine learning models. Specifically, the company plans to utilize these resources to enhance its search and recommendation systems. As a result, users will receive more personalized content suggestions based on their browsing habits.
Custom Processors for AI Workloads
The platform will adopt specialized hardware to run its large language models and vision-language models. Specifically, the company will use AWS Trainium chips to power its personalized visual search features. In addition, the company is expanding its use of Graviton processors, which currently run about one-third of its computing infrastructure.
“Pinterest is investing heavily in artificial intelligence to make discovery more personalized, visual, and actionable for hundreds of millions of our users every month. This expanded commitment with AWS gives us compute flexibility, hardware choice, and infrastructure efficiency.”
Matt Madrigal, Chief Technology Officer at Pinterest
Infrastructure Modernization with Kubernetes
The migration from traditional EC2 environments to a Kubernetes-based architecture represents a key technical shift. Notably, the platform will use Amazon Elastic Kubernetes Service to manage its containerized applications. This transition is expected to improve software development speed, operational reliability, and infrastructure efficiency across its global platform.
Long-Term Strategic Outlook
The Pinterest AWS deal represents a multi-billion dollar investment that ensures the platform can scale its generative AI models over the next five years. Furthermore, the integration of advanced machine learning tools will help advertisers improve their campaign performance. Ultimately, the collaboration aims to deliver faster search results and more accurate recommendations to global users.
“AWS is the ideal choice for AI applications at this massive scale, and we are committed to helping Pinterest teams move faster and think bigger.”
Dave Brown, Vice President of Compute and Machine Learning at AWS
