Drik Vision AI

Drik (Sanskrit: दृक् — "the seer") is a vehicle intelligence layer over any camera feed. Officers ask in plain language; Drik returns evidence frames, map, timeline, and graph. Seven atomic capabilities compose every scenario we ship.

We serve four verticals from one platform: Army and Defence formations on convoy, cantonment, and counter-intel work; SSB and border security on infiltration corridors and smuggling signatures; state and city police on post-incident backtraces, plate-cloning, and known-associate graphs; and traffic and municipal authorities on auto-challan deployments that pay for themselves from month one.

The technology is grounded in an in-house dataset of 500+ classes that covers the entire Indian-road long tail — from bullock carts to fire vans, from sugarcane jugaads to school buses — built by mixing traditional computer vision with modern large models. Facial recognition is being integrated. Beyond that, we are building toward the Drik Vision World Model: a vision foundation model that reasons about scenes the way a human brain does.