GeoSpy has evolved into something new. Introducing Raven, our frontline visual intelligence system. From a single image, Raven turns pixels into actionable intelligence in seconds. No metadata required.

Raven is using LGM - Large Geospatial Model. LGMs convert images, scans & 3D data into geospatial tokens - soil colour, shadow length, light angles. Unlike LLMs that process language, they process physical geometry. Raven doesn't need metadata.

There are lots of mechanical / car related auto-parts pictures just everywhere! especially structured in one parts, really useful data.

😉

there are like 100 of these platforms on the market now fyi

And only one of them works and is deployed internationally with law enforcement

credit @lillysharples for the image in the launch.

@graylark But still no lead on the 14+ missing NASA scientists…

Insanely cool, congrats

Thank you

Evil attack on privacy.

🕵️

This tool has no place in the public domain. It should be reserved for investigators and the police.

Luckily it is only for LE and approved enterprises

That is reassuring, in any case congratulations, this technology is truly amazing !

I blame the the jews

God's Eye ??

Getting disturbingly close to making this Red Dwarf scene reality:

Why not open source?

Will AI replace @georainbolt ?

Minority Report the sequel 🎥

🍆

Open source or bust

I'll try to upload my photo on it 😜

bro is cooking!

can I PLEASE get this for enterprise

pretty cool bro

This is sick. Have yall thought to compete with Cesium @heinenbros where we can run physics simulations on Isaac with topography/tiles to conduct mission planning?

How about demoin exact same feature but on some Pakistan picture. Locating a picture in SF is not that impressive tbh

Too bad you guys only sell your product to LE

cc @CL207

PHEW

Dan is a Fucking God

@GeospyAI just why

Hey @grok, fact check this post.

Awesome demo. crazy but awesome :)

Let's give it a try

insanity

Odin approved