Viso Secures $9.2M to Simplify Computer Vision for Enterprises
Computer vision technology is now prevalent in various industries, but developing and managing these advanced AI models remains a challenge. Viso is addressing this issue by offering a low/no-code end-to-end platform that empowers companies to build their own computer vision solutions. Recently, Viso raised $9.2 million to expand its operations.
While numerous computer vision models and services exist, many are often offered as “model as API” solutions. For instance, if you're looking to implement person recognition to assess whether individuals are standing or sitting, helping you gauge the busyness of locations such as train stations or restaurants, you face options that may not align with your specific needs, security standards, or budget.
Fully-formed person and pose recognition systems are available, but they might lack customization or be too costly for scaling. While building your own computer vision models is possible, the necessary expertise to train and deploy such models can be substantial. For most companies, assembling the right team with the skills and resources can be a daunting task.
Viso aims to solve this problem by providing a platform that enables enterprises to create high-quality computer vision models without the extensive commitment of time and resources typically required. "In the early stages of computer vision adoption, many companies rely on pre-built systems. Eventually, they realize the need to consolidate their computer vision initiatives, ensuring deep integration and customization. Moreover, data ownership is crucial due to its sensitivity and strategic value. This realization has led businesses to recruit AI engineers," explained Viso’s co-founder and co-CEO, Gaudenz Boesch.
Viso's Computer Vision Solutions
Unlike many enterprise needs, computer vision lacks a streamlined infrastructure for efficient development and deployment. "Organizations often find themselves piecing together various disconnected software and hardware systems, such as cameras and servers," Boesch noted. This disjointed approach demands cross-domain expertise, which can rapidly become costly.
Viso’s model will resonate with users familiar with no-code tools. The platform consists of a series of pre-built and customizable modules, allowing users to select, train, and deploy computer vision models effortlessly.
While some technical knowledge is still necessary—like determining which object recognition model to use, where to store training data, and how inference will be managed—a small team of engineers can operate much more efficiently. Everything can be managed in one consolidated platform rather than navigating multiple tools, APIs, and code repositories.
Viso promises an end-to-end solution, which is no exaggeration. Computer vision projects require data collection, training, implementation, hosting, and compliance work. Viso consolidates all these aspects into a single platform, offering a comprehensive solution.
For example, if you wanted to create a “busy detector,” you could potentially start with just a hundred hours of footage and emerge with a complete product in a week or two. This would encompass raw data analysis and storage, annotation and labeling, training and testing of your model, seamless integration into your product, deployment options (online or offline), analytics, ongoing updates, security, and access—all without leaving the Viso environment, likely without even writing a single line of code.
Although there are other platforms available, Boesch asserts that none are specifically designed to manage highly complex computer vision applications at scale while ensuring ongoing maintenance. Viso’s mission is to support a wide array of models, methods, hardware, and applications, allowing customers to retain ownership of their final product.
Even without being a developer, the appeal of using a comprehensive platform over several disjointed tools is evident, reflecting the rise of low-code and end-to-end solutions in various sectors.
Investors recognize this potential, and Viso successfully raised $9.2 million in seed funding, led by Accel and supplemented by strategic angel investors. Notably, Viso has been self-funded since its inception in Switzerland in 2018.
Boesch highlighted the increasing demand as the catalyst for this funding round, which, in the context of AI companies, is relatively modest considering the robust offerings and existing client base. Viso counts several large organizations, including PricewaterhouseCoopers, DHL, and Orange, among its clients and has seen a remarkable sixfold growth in new customers since 2022.
Ranking
