Apparat
12 posts

Apparat 已转推
Apparat 已转推

Same is true for video archives. The model gives you the intelligence; the context (how clips actually relate across an archive) is where the value is. That’s the whole bet at @ApparatSystems: give each archive its own brain.
Garry Tan@garrytan
I think one underestimated thing when we look back on it was how useful it is to have your own personal brain and company brain in 2026 at the dawn of usable AGI AGI gives you the intelligence You still have to collect your personal context to get the real unlock
English
Apparat 已转推

Another Apparat update:
Been working on making our query agent understand creative briefs.
Give it a promo script or highlight reel brief, and it breaks that into shot requirements before searching the archive for matching footage.
Definitely one of the more interesting problems we’ve worked on so far.
English

The intelligence layer for video archives
Dylan Romay@DylanRomay_
One thing we've been working on before launching Apparat: Teaching our agents that not all video archives are the same. The way you analyze a football match, a documentary, or a news broadcast is fundamentally different. The agents need to reason with the right context.
English
Apparat 已转推

Our landing page for @ApparatSystems is live!!!
A video intelligence platform that indexes and searches large video archives!
Check it out, would love to know what you think!
apparatsystems.com
English
Apparat 已转推
Apparat 已转推

and less than 2% of all video generated globally is actually indexed or actively analyzed
Dylan Romay@DylanRomay_
Did you know that 4.1 billion hours of video are generated every day? This is equivilent to ~1 billion movies. Video represents 90% of the world's digital data
English
Apparat 已转推
