LGND wants to make Chatgpt for the Earth

The Earth is mired in the data itself. Every day, capture satellites around them 100 terabytes of pictures.

But his understanding is not always easy. Simple questions can be very complicated to answer them. Take this question from Vital economic importance to California: How many fire breaks have a country that has stopped the wildfire in its paths, and how has changed since the last firefight season?

“Originally, you will have someone looking at the pictures. This is hesitant so far,” Nathaniel Manning, co -founder and CEO of the company LGNDTell Techcrunch. In recent years, nerve networks have made somewhat easier, allowing automatic learning experts and data scientists to train algorithms on how to see firewalls in satellite images.

He said: “It is possible that you know, as you know, a few hundred thousand dollars – if not several hundreds of a thousand dollars – to try to create this data set, and it will only be able to do so one thing.”

LGND wants to lower these numbers in order in size or more.

“We are not looking to replace the people who do these things,” said Bruno Sanchez Andradi Nuno, co -founder of senior scientists at LGND. “We are looking to make it 10 times more efficient, and a hundred times more efficient.”

LGND has recently raised a $ 9 million seed tour led by Jaffelin Venture Partners, and the company told Techcrunch exclusively. Share Aenu, ClockTower Ventures, coalition operators, MCJ, OverTure, Ridgeline and Space Capital. Also joined a number of owners investors, including Keyhole John Hanke, co -founder of Ramp Karim Atiyeh, and Salsforce Suzanne Dibianca.

The primary product for starting is a geographical data vector guarantees. Today, most geographical information is found in pixels or traditional vectors (points, lines and regions). It is flexible, easy to distribute and read, but the interpretation of this information requires either a profound understanding of the space, or some of the non -trivial quantity of computing, or both.

Geographical implications summarize spatial data in a way that makes it easy to find relationships between different points on the ground.

Nonu said: “The implications get 90 % of all the unique accounts in the foreground,” Nonu said. “The implications are high -end global summaries that embody 90 % of the account that you should do anyway.”

Take an example of fire breaks. They may take the shape of roads, rivers or lakes. Each of them will appear differently on the map, but they all share certain properties. For one, the pixel units that make up a picture of a fire break will not have any plants. Also, the fire break should be a certain minimum, which is often dependent on the length of the vegetation around it. The implications make it easy to find places on the map that matches these descriptions.

LGND has built an institution application to help large companies answer questions that include spatial data in addition to the application programming interface that users with specific needs can arrive directly.

Manning sees LGND contents to encourage companies to inquire about spatial geographical data in completely new ways.

Imagine Amnesty International’s travel agent. Users may ask to find a short -term rental with three rooms close to good diving. “But also, I want to be on a white sandy beach. I want to know that there is a few seaweed in February, when we go, and perhaps more importantly, at this time of reservation, there is no one kilometer at a distance from our house,” he said.

Building traditional spatial geographical models to answer these questions will be a waste of time for only one inquiry, not to mention all of them together.

If LGND can succeed in providing such a tool to the masses, or even for people who use spatial geographical data for their functions, then it has the ability to take a bite of the market Near 400 billion dollars.

“We are trying to be the standard oil of this data,” said Manning.