Despite the widespread use of computer vision in many industries, developing and managing these visual AI models is still challenging. Viso Eyes is developing an end-to-end low-code/no-code platform that enables businesses to develop their computer vision stacks, and they recently raised $9.2 million to expand.
Indeed, there are many computer vision models and services available, but many of them can be loosely classified as “model as API.” Let’s say you wish to identify people and determine if they are standing or sitting to gauge the volume of activity at a restaurant or train station.
You can find fully-formed choices for posture and person recognition, but they might need to match your security model or use case or be too costly to expand. While it is possible to build your own, doing so may be beyond your means Viso Eyes unless you have the resources and time to establish a professional team. Modern CV models require a high level of skill to train and implement.
Viso aims to address situations like that by allowing you to design your enterprise-grade CV model without investing the time and resources often required.
“In the early stages of the adoption cycle, businesses purchase or rent pre-made computer vision equipment. All computer vision initiatives must eventually be streamlined, thoroughly integrated, customized, and “owned” because the technology is strategically valuable and the data is sensitive. This is why businesses in those sectors are beginning to employ AI engineers, according to Gaudenz Boesch, co-founder and co-CEO of Viso.
However, unlike many other enterprise-level requirements, computer vision lacks the “specialized infrastructure” to design and deploy it effectively.
Viso Eyes No-Code for the Future of Computer Vision:
Viso Eyes No-Code for the Future of Computer Vision [Source of Image: Techcrunch.com]
“Enterprises must construct it from the ground up, attempting to put together a multitude of disjointed software and hardware platforms (cameras, servers) throughout the company,” he said. This, in turn, necessitates multidisciplinary competence, which quickly becomes prohibitively expensive.
Anyone who has used no-code tools in other contexts would probably recognize Viso’s approach. “It consists of a collection of Viso Eyes pre-configured and customizable modules that empower users to select, train, and utilize computer vision models based on their needs.
Compared to other available computer vision platforms, Boesch asserts that none were designed to efficiently handle intricate, large-scale computer vision applications and provide continuous maintenance. Instead, they primarily focused Viso Eyes on a limited set of functionalities mentioned earlier. Viso’s goal is to ensure that the end-user owns the final product while accommodating a wide range of models, techniques, hardware, and use cases.”
Investors in Viso appear to agree, as the startup has raised $9.2 million in seed funding led by Accel, including participation from several angel investors. Remarkably, the business was bootstrapped and established in Switzerland in 2018.
According to Boesch, the company had to hike due to the rapidly increasing demand, even though the raise is minor compared to AI’s Viso Eyes products and clientele already in place. He noted that since 2022, Viso has added six times as many new customers as existing ones. Several major corporations, like PricewaterhouseCooper, DHL, and Orange, have already embraced the software.