DRIK REASONING ENGINE

Visual Reasoning AI

The world has a billion cameras. None of them can reason about what they see.

Until now.

Scroll

Drik Vision AI

Visual intelligence for every camera on earth.

A vehicle intelligence layer over any camera feed. Officers ask in plain language — Drik returns evidence frames, map, timeline, and graph. No GPU upgrades. No retraining. Just plug in and ask.

Seven Atomic Capabilities

Every scenario is a composition of these primitives.

Officers ask natural queries. Drik returns evidence frames, map, timeline, and graph. The same seven primitives serve army, border, police, and traffic.

  1. 01

    Vehicle of Interest Trace

    Every historical sighting across every feed.

  2. 02

    Predicted Current Location Cone

    From last sighting + direction + speed.

  3. 03

    Predicted Home Area

    Dominant overnight zone.

  4. 04

    Habitual Route Inference

    Repeated-path corridor.

  5. 05

    Co-travel Graph

    Vehicles that repeatedly co-appear.

  6. 06

    Visual-Description Search

    "White Bolero, dented bumper, roof carrier" — ranked results.

  7. 07

    Auto-Challan on Violations

    Helmet-less, red-light, wrong-way, overspeed (cross-camera), triple-ride, seatbelt, phone, tint, plate tampering, no-parking, school-zone.

Capability 7 is the self-funding wedge. Capabilities 1–6 are the intelligence depth.

Four Verticals

One platform. Four buyers.

The seven atomic capabilities compose into 18 deployment scenarios across army, border, police, and traffic.

Army / Defence

Scenarios 1–6

Convoy, cantonment, and counter-intel work that a radio race cannot keep up with.

  • Vehicle of Interest Trace across every ingested feed
  • Predicted current location cone for QRT dispatch
  • Pattern-of-life profile: dwell heatmap, dominant night-zone
  • Co-travel / convoy graph from temporal co-appearance
  • Dwell-near-installation alerts with no officer in the loop
  • Cross-formation hand-off when the target crosses AOR

SSB / Border Security

Scenarios 7–9

Infiltration corridors, handler–asset relationships, and behavioural signatures of smuggling.

  • Border-corridor recurrence: foreign-origin or repeated approach-and-retreat patterns
  • Handler–asset correlation: vehicle pairs co-appearing only at drop zones
  • Smuggling behavioural signature: slow night approaches, unlit vehicles in sensitive sectors

Police — Investigation & Law-and-Order

Scenarios 10–15

72 hours of 12 feeds compressed to minutes. Network discovery without prior intel.

  • Post-incident backtrace inside a spatiotemporal cone
  • Stolen vehicle re-identified through physical signature, even with swapped plates
  • Plate-cloning detection across cities and same-day windows
  • Known-associate graph from a suspect’s vehicle co-travel
  • VIP-route tail detection: shadowing flagged before the event
  • Beat-level anomaly: density baselines surface out-of-area clustering

Traffic + Municipal — Self-Funding

Scenarios 16–18

Zero capex to the state. Majority share of challan revenue stays with the buyer.

  • Auto-challan across 11 violation classes (helmet, seatbelt, red-light, wrong-way, triple-ride, phone, overspeed, parking, silencer, tint, school-zone)
  • Fake-plate detection where text disagrees with physical signature
  • Revenue-share deployment: the system pays for itself from month one
The system pays for itself from month one — the deployment economics that fund the rest of the platform.

Selected Scenarios

Four scenarios in motion

The self-funding wedge first, then three intelligence-depth scenarios. Each composes one or more of the seven atomic capabilities above.

Camera feed Camera feed
CAM-03 · LIVE · 30fps
ENFORCEMENT MODE
Begumpet Rd
Capability 7 · Auto-Challan · Self-Funding Wedge

The wedge that pays for the rest.

Eleven violation classes. Zero capex to the state. Majority share of revenue stays with the buyer — the system pays for itself from month one and funds the intelligence depth that follows.

Total Revenue Generated
₹0
LIVE CHALLANS
0
Challans
0
Plates
Camera network map
Capability 6 · Visual-Description Search

Describe anything. Find it everywhere.

SEARCHING 847 CAMERAS
MATCH · 94.7% CAM-009 · Pal Road · 16:28:41
Decorated Indian truck
Decorated Truck · "Blow Horn" · RJ-14-TB-8834
"Mera Bharat Mahaan" · Lotus · Red swastik hangings
MATCH · 87.2% CAM-004 · 16:14:22
Decorated Indian truck
"Horn Please" · RJ-27-GA-3045
Hand-painted rural scenes
MATCH · 72.1% CAM-012 · 15:48:09
Decorated Indian truck
"Blow Horn" (Punjabi) · PB-10-DK-8645
Painted landscape art · Sarbat da bhala
Capability 1 · Vehicle of Interest Trace

One image. Every sighting. Every camera.

DROP IMAGE
Source
EXTRACTING VISUAL FEATURES...
96.3%
CAM-009 · 16:28
91.7%
CAM-004 · 16:14
78.4%
CAM-012 · 15:48
74.1%
CAM-003 · 15:22
62.8%
CAM-011 · 14:55
58.2%
CAM-006 · 14:31
Capabilities 2 + 5 · Predict + Co-travel

Watch. Predict. Surface convoys.

ACTIVE WATCHLIST

Flag any vehicle or person of interest. The system watches every camera feed, 24/7.

RJ-14-XX-9999
JUST SPOTTED
LIVE ALERTS

The moment a watched target appears on any camera, you get an instant alert with location and confidence.

RJ-14-XX-9999
CAM-012 · Shastri Nagar · 14:32:07
Match confidence: 97.2%
ROUTE PREDICTION

See where a vehicle is heading before it arrives. Predicted routes from historical patterns.

Predicted next Ratanada Circle
ETA ~12 minutes
Confidence
78%

In-house Dataset

500+ classes. The Indian road, in full.

Bullock carts to Boleros, fire vans to e-rickshaws, sugarcane jugaads to school buses — trained on the messiest road network on earth, not on COCO.

Cars

  • All brands
  • Make + model + year
  • Colour + customisation
  • Roof carriers, bull bars, body wraps

Two-wheelers

  • Motorcycles
  • Scooters (electric + ICE)
  • Mopeds
  • Rider + pillion + load configurations

Commercial — Goods

  • Tata Ace, Bolero Maxi, Ape
  • Medium + heavy trucks (Tata, Ashok Leyland, BharatBenz)
  • Tankers, container trucks, low-bed trailers
  • Pickup trucks (formal + informal)

Public Transport

  • State + private buses
  • School + sleeper + semi-sleeper
  • Mini-buses
  • Auto-rickshaws (petrol/CNG/electric)
  • Tempo travellers

Emergency

  • Ambulances (108-series, private, basic)
  • Fire vans
  • Police gypsies + prisoner vans
  • Disaster response vehicles

Agricultural

  • Tractors (all major brands)
  • Tractor-trolleys (farming, construction, municipal use)
  • Combine harvesters
  • Pump sets on trolleys

Municipal

  • Garbage trucks (compactor + open)
  • Garbage tractor-trolleys
  • Water + sewage tankers
  • Street sweepers, bucket trucks

Informal / Jugaad

  • Sugarcane juice jugaads
  • Mobile food carts (motorised + pedal)
  • Cargo e-rickshaws
  • Modified-chassis transport
  • Improvised passenger vehicles

Animal-drawn

  • Bullock carts
  • Horse carts (urban tongas, rural)
  • Donkey + camel carts (regional)

Human-powered

  • Cycle rickshaws (passenger)
  • Hand-pulled cargo carts
  • Tricycle vendors

Construction

  • JCBs / backhoe loaders
  • Excavators
  • Concrete mixers, dumpers
  • Mobile cranes
  • Aggregate-haul tractor-trolleys

Built by mixing traditional computer vision with modern large models. Pure LLM pipelines are too slow at camera-feed scale; pure classical detectors miss the long tail. Drik runs classical perception at the edge and uses large models surgically — for description, refinement, and ambiguity resolution.

How It Works

Your cameras. Our intelligence. Any scale.

Deploy Drik on-premise, in the cloud, or both. Our nodes scale with your camera network — from a single building to an entire city.

Camera network infrastructure
CAM-001
CAM-002
CAM-003
CAM-004
CAM-005
CAM-006
CAM-007
CAM-008
CAM-009
CAM-010
CAM-011
CAM-012
DRIK NODE
< 17ms Full Latency
4x RTX PRO 5000 GPU Cluster
200+ Cameras
On-Premises Deployment

Five Levels of Visual Reasoning

From pixels to understanding

Level 1

Detect

It sees every object

Every vehicle, every person, every object — detected, classified, and counted in real time across every frame.

Level 2

Recognize

It reads every plate, every face

License plates extracted. Vehicle makes and models identified. Colors, modifications, damage — all catalogued instantly.

Level 3

Describe

It describes what it sees in words

Not just labels. Full natural-language descriptions of every scene, every event, every anomaly. Searchable. Queryable.

Level 4

Reason

It understands cause and effect

Why did traffic stop? What caused the accident? Which vehicle triggered the chain reaction? The AI builds causal chains.

Level 5

Predict

It anticipates what happens next

Pattern recognition across time and space. Congestion forecasts. Accident risk zones. Before it happens, the system sees it forming.

What’s Next

Drik Vision World Model

Today: vehicle intelligence over any camera feed. Next: a vision foundation model that reasons about scenes the way a human brain does — not frame-by-frame detection plus classification, but a unified system that builds and updates a model of the world from camera feeds.

Early R&D · Roadmap pillar

DVWM is research, not product. No demo, no benchmarks. The honest framing: we are building toward it because the next decade of camera AI is reasoning, and per-scenario engineering does not scale to it.

Indian traffic

The Proving Ground

We chose the hardest visual environment on earth.

1.4 billion people. 300 million vehicles. Zero lane discipline.

Jugaads. Ox carts. Overloaded trucks. Band baarats. Cows.
Modified vehicles. No lanes. No structure.

If it reasons here, it reasons everywhere.

50+
Vehicle Types
12
Indian States
24/7
All Conditions
Real-time
Processing

The People Behind Drik

Built by those who refuse to accept blind cameras

Hansraj Patel

Hansraj Patel

Founder & CEO

IIT Guwahati
Forbes India 30 Under 30 · 2025

Forbes India 30 Under 30 (2025). IIT Guwahati alumnus building what cameras were always meant to be — systems that truly see, reason, and act.

"India has 50 million CCTV cameras and almost none of them think. I want to change that."

Track Record

2020–2023
Full-cycle entrepreneurial experience — fundraising, hiring, product development & go-to-market

Backed By

Forbes India 30 Under 30

2025 · Recognized among India's top young innovators

TIDF · IIT Guwahati

Incubated & backed by DST, Govt of India

Govt of Assam

Backed by the State Government

Industry Traction

LoIs from government offices · ITC & system integrator partnerships

The Team

Hansraj Patel

Hansraj Patel

Founder & CEO

+

We're hiring

Computer Vision Engineer

Full-time · Remote · India

Apply →

Full-Stack Engineer

Full-time · Remote · India

Apply →
View all open roles →

See what your cameras are missing.

Built in India. For the world.