An interview conducted by MONOist with Tiffany Bloomquist, Head of Startups, Asia Pacific and Japan at Amazon Web Services, and Taesan Kwon, Chief Operating Officer at STRADVISION, explains how cloud infrastructure and startup support programs can help a company specializing in computer vision move from developing specialized technology to deploying AI solutions across multiple industries.
The interview took place during a media event organized by AWS at STRADVISION’s office in South Korea on September 8, 2026. It focused on the two companies’ work in AI and on what AWS calls physical AI: the use of intelligent models in environments that interact with the real world, such as vehicles and robots.
From Idea to Market
Bloomquist said that AWS has invested more than $80 billion over 20 years and supports more than 350,000 startups worldwide through the AWS Activate program. According to her explanation, this support is not limited to computing credits and technical services; it also extends to helping companies build products, develop go-to-market strategies, use AWS Marketplace, and connect with enterprise partners and customers.
AWS presents companies such as Netflix and Grab as examples of businesses that began with specific ideas and then expanded into global operations. However, the interview provides no new figures on STRADVISION’s growth or specific commercial results; instead, it focuses on the mechanisms AWS uses to support companies during development and expansion.
What Does AWS Provide for AI Solutions?
AWS says it provides more than 200 services that can be combined according to customer needs. For compute-intensive AI projects, options include using NVIDIA graphics processing units or accessing AWS Trainium chips for model training and AWS Inferentia for running inference.
The company also referred to AWS SageMaker HyperPod for coordinating large-scale model-training operations and Amazon Bedrock for accessing AI models through application programming interfaces, in addition to Amazon Q Developer and Kiro for software development work, and Amazon QuickSight for analyzing and presenting data.
STRADVISION’s Applications in Physical AI
STRADVISION develops AI-based computer vision solutions and uses them in vehicle-related applications. Bloomquist said that the company’s technologies represent an early use case for physical AI, while models trained on automotive data are being expanded into other fields, including robotics.
According to the article, STRADVISION uses AWS solutions in developing its products. However, the interview does not specify which services are used at each stage of production, nor does it disclose performance indicators, customer contracts, or technical details that would make it possible to independently measure the impact of these services.
Why Does This Development Matter?
The significance of this case lies in showing that developing AI for real-world environments depends not only on the model itself, but also on infrastructure for training and running models, as well as tools for data management, development, and commercial distribution. For startups in the automotive and robotics fields, this integration may reduce the need to build the entire infrastructure internally.
At the same time, the success of this model remains linked to data quality, safety and reliability requirements in mission-critical applications, computing costs, and the company’s ability to turn a prototype model into a deployable product. The article does not provide enough detail to assess these aspects in STRADVISION’s case. Therefore, the interview should be viewed as a presentation of a technology partnership and support ecosystem, not as an independent evaluation of product performance.