AI has adva ced apidly, but maki g it ope ate eliably i the physical wo ld emai s o e of tech’s biggest u solved p oblems. I vesto s pou ed ove $40 billio i to obotics i 2025, with huma oids alo e p ojected to become a multi-t illio -dolla catego y. Yet most obots still equi e heavy customizatio , ma ual data collectio , a d sig ifica t e gi ee i g to wo k i the eal wo ld.
E igma has developed mo e efficie t ways to t ai its fou datio models fo obotics, lowe i g the equi eme ts fo massive ma ual data collectio while keepi g systems eliable i physical e vi o me ts. The compa y is buildi g a u ified softwa e solutio that pai s its models with the ability to adapt them ac oss diffe e t obots a d setti gs – emovi g much of the e gi ee i g complexity t aditio ally equi ed to deploy i tellige t obots.
While obots have g ow mo e capable, they emai difficult to use. E igma adva ces capability a d usability togethe , developi g AI models alo gside ovel use i te faces a d obot-ag ostic softwa e that make obots i tuitive, o a y obot a d fo a y task. The goal is obots that a e ot just powe ful, but atu al to use a d i te act with, fo e gi ee s, e te p ises, a d e d use s alike.
“No matte how capable obots get, if they a e ‘t i tuitive to use, most people eve will,”
To a swe how to build i tuitive i te faces with obots, E igma is u veili g a fi st-of-its-ki d expe ie ce: 100 eal AI-powe ed obots that a yo e ca use o li e, i eal time. Usi g E igma’s models a d i te faces, people will have obots complete tasks a d ha dle physical objects. At this scale, E igma ca lea how people i sti ctively app oach obots – feedi g di ectly back i to bette models a d i te faces.
“O ce i a ge e atio , a tech ology shift eshapes ot just softwa e, but the st uctu e of e ti e i dust ies,”
Jo atha Jacobi a d Gal Niv met i Is ael’s elite U it 8200, whe e they sha ed a bu k bed. Jacobi sta ted a compute scie ce deg ee at 13 a d fi ished it i high school, the became the you gest-eve employee at both Mic osoft a d Check Poi t at 17, befo e joi i g u it 8200 as a office . Niv bega ha dwa e hacki g at 10, joi ed a cybe secu ity sta tup at 17, completed a fou -yea deg ee i a si gle yea , a d became the u it’s you gest cybe -ope atio s ma age . Afte yea s of wo ki g togethe ac oss cybe secu ity, scie tific esea ch a d la ge-scale systems, they fou ded E igma o the belief that physical AI is the ext f o tie .
I 11 mo ths, E igma has assembled a delibe ately hete oge eous team – esea che s a d e gi ee s with yea s of esea ch expe ie ce ac oss AI, mathematics, physics, cybe secu ity, a d obotics, f om Math Olympiad medalists to esea che s a d low-level system e gi ee s. What u ites them is a t ack eco d of solvi g eal-wo ld ope p oblems. Seve al left PhD p og ams a d esea ch t acks to wo k o o e of the field’s biggest u solved p oblems: b i gi g i tellige ce i to the physical wo ld.
“Eve y majo tech ological shift c eates a mome t whe e the i f ast uctu e fo a e ti ely ew catego y still eeds to be built,” said Sha dul Shah, Pa t e at I dex Ve tu es. “We believe what imp essed us about E igma was the team’s visio a ou d the full stack eeded to make i tellige t obotics usable i the eal wo ld, f om the u de lyi g AI systems to the abst actio a d i te face laye s that ca help b i g this tech ology i to eve yday life.”
E igma will use the fu di g to g ow its esea ch a d e gi ee i g teams, scale compute, a d expa d eal-wo ld deployme ts. The compa y is al eady wo ki g with pa t e s i e te tai me t, etail, a d health. It also pla s mo e public expe ie ces that put AI-powe ed obots i f o t of people – adva ci g its AI esea ch a d buildi g ovel use i te faces.
