Making AI accessible to every enterprise

In the 1990s, the business world underwent a seismic shift with the advent of the internet. It changed the way companies did ever y thing from creating their products to selling them to consumers. These days, companies are poised on the edge of another massive transition toward artificial intelligence and machine learning. “AI is going to touch every company, in every industry, in a material way,” says Rodrigo Liang, cofounder and CEO of AI solutions provider SambaNova. “It’s going to be such a competitive advantage.”

Yet, AI and machine learning isn’t accessible to every company. In-house AI systems typically require large staffs and heavy investments in infrastructure, putting the technology out of reach for many organizations. Liang and SambaNova co-founders Kunle Olukotun and Christopher Ré rejected that resource-intensive approach, aiming instead to democratize AI with an extensible subscription-based machine-learning service platform that doesn’t require an army of experts to make it work. “Some of the top companies have the ability to establish AI labs with hundreds and hundreds of AI scientists and data scientists,” Liang says. “But the large majority of companies in the world aren’t doing that.”

SambaNova- for letting companies train AI models in the cloud

The platform is designed to help organizations efficiently— and cost effectively—tap into the power of AI and machine learning to enable natural language processing, high-resolution imaging, and powerful recommendation. SambaNova’s Dataflow-as-a-Service helped earn SambaNova a spot on Fast Company’s 2021 list of the Next Big Things in Tech.


SambaNova began as a research project at Stanford University. Liang and his SambaNova cofounders were looking for ways that companies without in-house AI specialists could gain insights from the increasingly vast amounts of data available to them. The company took a software-first approach, initially designing a system optimized for managing data. They then created the hardware needed to optimally run that application, going so far as building their own specialized microprocessors.

SambaNova’s from-the-ground-up mindset was instrumental in helping it devise a novel solution to the challenge of efficiently delivering AI and machine learning capabilities. “We don’t have legacy architectures, legacy interfaces, or legacy constraints that keep us from solving problems the right way,” Liang says.


AI-as-a-Service helps provide organizations with the tools they need to make important innovations of their own. For example, cancer researchers have partnered with SambaNova to train AI to detect cancer cells in ultra-high-resolution images. The AI technology is far more sensitive than the human eye and, in some cases, is able to draw a boundary around individual areas of cancer cells.

Technology like this has broad applications. Once this system is up and running, healthcare workers can deploy it in other parts of the world, bringing cancer-detection services to communities that would otherwise have no access. Researchers are using the technology to help discover COVID-19 treatments, to perform defect analysis for manufacturing, and to identify areas of dark matter in the universe. “It’s amazing to work with expert partners who have done this type of work for many years,” Liang says. “We’re able to come in, and in a very short amount of time, give them a big boost in their capabilities.”