Enterprise automation company Landing.ai has hired Amazon director of artificial intelligence and computer vision Gopi Prashanth to act as its new VP of engineering. Prashanth led a team that created the AI system behind Amazon Go, and previously worked as a development manager at Microsoft.

Prashanth is one of the biggest hires yet for Landing.ai, a company founded in December 2017 by Google Brain cofounder and former Baidu AI chief scientist Andrew Ng.

Ng is also cofounder of online education company Coursera and the $175 million AI Fund to back promising startups.

“I think Gopi is one of the unsung heroes of AI,” Ng said. “I think Amazon Go is one of the most amazing technologies to land in the AI world in the modern era, so Gopi kind of built the core AI technology that made Amazon Go possible. He built AI for one of the most amazing products in our generation. Before that he led Project Catapult at Microsoft which wound up helping transform how Microsoft uses FPGAs in its cloud GPU.”

In his new role, Prashanth will build the Landing.ai engineering team, expand its ventures into more industry verticals, and meet a growing demand for novel applications of AI. He will also oversee the company’s software development and applied science teams, and set the agenda for research initiatives.

Landing.ai has generally been secretive about the businesses and organizations it works with, but publicly announced deals include work with Foxconn and in recent weeks a partnership with LG.

Since launch, Landing.ai has focused primarily on manufacturing automation, but has since ventured into working with customers in auto and silicon chip manufacturing, Ng said.

Many of the AI systems the company makes available to customers today include computer vision, like Landing Light for visual product inspections and other tasks. Landing.ai will continue to focus on providing AI models for businesses that can produce reliable results with small amounts of data, something Ng told VentureBeat was a challenge AI companies will face in 2019.

“A lot of the focus of Landing.ai has been to build a unique tech stack that works even on small data, and this tech stack we’re applying definitely to manufacturing, but also other problems in agriculture and health care, for example, where you may not have a million x-ray images in a specific position but much lower datasets. These tools allow our algorithms to be applicable even with more modest amounts of data,” Ng said.

As part of its strategy, Landing.ai plans to work with typically underserved industries where AI can add value, create new products, improve existing ones, or gain efficiency. The focus will also be to capture and automate tasks that typically take a human no more than one second to complete, and transfer what Ng and Prashanth refer to as tribal knowledge.

Going forward, Prashanth sees opportunities ahead that can arise from the use of geospatial satellite imagery and computer vision to find everything from oil wells to droughts.

“Being able to do this in an automated manner will allow for better living conditions for all of us so we can actually influence environmental change using AI. So these are some areas where today, if you think hard enough, there are things that we can do that make a better life and environment for us,” Prashanth said.

Landing.ai currently has roughly 50 employees and is based in Palo Alto, California.