★ Selected — IRF 2026 Start-Up Label · 1 of 5 worldwide, the only one from Asia · Geneva, Nov 2026
Incubated at AIC IIT Delhi  ·  Live fleet pilot

Every vehicle, a road-safety sensor

Civilytix turns the dashcams and phones already in vehicles into a real-time road-hazard network — warning drivers in Hindi in under a tenth of a second, even offline, and sharing every hazard to a live map the whole fleet can see.

Partner with us See the live data ↓
10,688 km surveyed  ·  1,408 hazards graded  ·  Live on iOS & Android
Backed & recognised by
<0.1 s
On-device Hindi alert
16-category
Hazard detection
Offline
Works without network
Zero
New hardware
V2V / V2X
Shared live map
Global recognition

On the world stage for road safety

IRF
2026 Label

Selected for the IRF 2026 Start-Up Label

The International Road Federation named Civilytix one of five start-ups worldwide — and the only one from Asia — in its 2026 Start-Up Label, cited for combining edge-AI hazard detection with live road-risk maps for safer mobility in the Global South.

🌍 Pitching at the IRF Annual Conference 📍 Geneva · 16–18 Nov 2026 🇮🇳 Only finalist from Asia
Traction

Already on the road — real kilometres, real hazards

Running today with a paying truck-fleet pilot in Faridabad. Every drive adds road to the map and sharpens the dataset.

0
of road network surveyed by the fleet
0
road hazards graded from real footage
0
of Indian road footage analysed
0
hazard categories detected & mapped
Paying fleet pilot · Faridabad Live on iOS & Android Privacy-first · faces & plates auto-blurred Hosted in India · DPDP-aligned
The data, not a demo

What a road actually looks like, if you count

Every dot on our map is a real detection with its own photograph, GPS coordinate, timestamp and severity grade. Here's a live-style sample from the fleet, and the real breakdown of the first 1,408 hazards.

Detection feed

Live sample

First 1,408 hazards · by type

Only three potholes. On a corridor drivers call dangerous — because on a good concrete highway the danger isn't the surface. It's the 649 vehicles stopped where they shouldn't be and the 519 places the lane markings have worn away, on roads where India mandates lane-departure warning from 2028. We listen to the data, not the assumptions.
What we're building

A road-safety network on the cameras vehicles already have

On-device AI spots hazards — potholes, barricades, debris, stopped vehicles, pedestrians, lane drift — and warns the driver in Hindi in under a tenth of a second, even offline. Each hazard is shared vehicle-to-vehicle so the truck behind knows too. Underneath, it builds a road-safety dataset for Indian roads.

How we're honest about it: we integrate proven camera hardware (ADAS/DMS + dual recording) and build the network, app, Hindi alerts, privacy layer, insurance-evidence workflow and fleet dashboard on top. That's our value — we're not claiming we built the camera chip.
In your pocket

The Civilytix app — live on iOS & Android

Hazard warnings spoken in Hindi, a live map the whole fleet shares, travel modes tuned to your vehicle, and privacy built in — no login, faces and plates blurred on your own device.

Civilytix app — live hazard map with navigation and dashcam feed
Civilytix app — choosing a travel mode: car, motorcycle, bicycle or pedestrian
Civilytix app — privacy consent screen explaining on-device anonymisation
Live map · Travel modes · Privacy consent
How it works

Capture, detect, map — in three steps

STEP 01

Capture

The dashcam or phone already in the vehicle streams the road ahead — no new hardware, works on fleets you already run.

STEP 02

Detect & alert

AI flags hazards on the road and warns the driver in Hindi in under 0.1 s — on the device, even with no network.

STEP 03

Map & share

Confirmed by multiple vehicles, each hazard lands on a live map and is shared V2V/V2X — a safety layer every connected driver and city can use.

Our vision

Toward zero road deaths — the "Safe Swarm"

Built for the vehicle that actually kills

Two-wheelers account for 28.4% of pedestrian deaths — more than cars. Western ADAS was designed for a car on a lane-marked road. We build for the other case: the rider, the truck, the unmarked Indian road.

A network that compounds

Every additional vehicle makes the hazard map richer and the dataset stronger for everyone — a public good and a durable competitive advantage.

Why now

The need and the regulation are arriving together

1,72,890road deaths in India in 2023 — an all-time high (MoRTH)
~1 in 5people killed on Indian roads is a pedestrian
28.4%of pedestrian deaths involve two-wheelers — more than cars
2028lane-departure warning set to be required on Indian highways
Who it's for

Practical impact, for everyone on the road

StakeholderWhat they get
Drivers
truck / commercial, gig 2-wheelers
Real-time Hindi warnings that help prevent crashes (hazard, collision, lane, pedestrian); fatigue/phone nudges on long hauls. Video protects them from false blame. Good driving shows up in a safety score. It's on your side — not surveillance.
Fleet ownersEvery truck, driver and route on one live screen. Fewer accidents mean fewer write-offs, less downtime and, over time, better insurance terms. Driver accountability (fatigue/phone/belt) without sitting in the cab. Geo-fencing for theft/deviation alerts. Compliance-ready as norms tighten.
Logistics operatorsRoute-risk intelligence for safer, more predictable ETAs; fewer delays from accidents and breakdowns; cargo/vehicle security via live tracking; a road-condition dataset that improves routing over time.
InsurersVerifiable, timestamped, geotagged incident evidence plus road-risk data — for sharper pricing, faster claims and less fraud. Our evidence helps support claims; it never guarantees a payout.
Cities & road authorities
B2G
A live pothole/hazard map for data-driven repair prioritisation — we're already offering this dataset to NHAI and state road authorities. Human-in-the-loop and privacy-first: faces and plates blurred on-device, DPDP-aligned.
OEMs / map & navigationA real-time in-cab hazard layer built for Indian roads — differentiated ADAS and navigation, with no new hardware.
Vulnerable road users
2-wheeler riders, pedestrians
The people who die most. Hazard and blind-spot alerts target the most lethal collision types — the core of the mission.
Society / the nationMovement toward zero road deaths, plus a road-safety dataset for Indian roads — a public good and our competitive moat.
In one line

The same mission, in each audience's words

Driver
"यह आपके साथ है — warns you, protects you."
Fleet owner
"Fewer crashes, cheaper insurance, drivers accountable — one screen."
Insurer
"Proof and road-risk data that cut fraud and speed claims."
Government
"A live hazard map that tells you which roads to fix first — privacy-first, the official stays in charge."
Investor
"Road-safety AI plus a compounding India-first data network, on hardware fleets already own."
Team

The founders

HS

Harshit Sharma

Co-Founder & CEO
Computer Vision · Edge AI · ADAS. Road-quality research at IIT Kharagpur; national winner, DTU Skylabs (Denmark). Deep-tech for Bharat.
RS

Ria Singhal

Co-Founder & COO
Civil Engineer. Scaling the Hazard Map & Blindspot platform and the Faridabad fleet pilot. Deep-tech for Bharat.
Contact

Let's build safer roads together

Fleet operator, insurer, OEM, city, or investor — tell us a bit and we'll get back within 24 hours.

Prefer email? harshit.sharma@civilytix.com  ·  Investors: request data room