India-Origin AI · C3iHUB · IIT Kanpur

Sovereign AI platforms that detect, identify, and respond to threats — on the ground and in the air.

Shatakshi counter-UAS detection, an Anti-Drone VR Simulator, and SkyEye AI monitoring — an indigenous airspace and perimeter security platform built with C3iHUB at IIT Kanpur, field-tested with Army units, and priced for mass deployment.

Backed By · Validated By

C3iHUB · IIT Kanpur
Field-Tested with Army Units
Patent-planned Detection Architecture
Atmanirbhar Bharat Aligned

Two products, one self-reinforcing platform
— plus AI monitoring for the ground.

Shatakshi detects and neutralises aerial threats. The Anti-Drone Simulator trains operators on the exact same sensor logic, with zero live-drone risk. SkyEye extends AI vision to your ground-level camera network.

Shatakshi

Counter-UAS detection system — EO/IR + RF sensor fusion

8-10 fixed EO/IR cameras + PTZ, 1-1.5km day detection, <200ms latency. Entry price under ₹50 Lakh. Field-tested with Army units.

Explore Shatakshi

Anti-Drone Simulator

Physics-true VR training — detection in the loop

Unreal Engine 5, real sensor physics, adaptive swarms up to 50 drones. The only trainer that uses the same sensor logic operators will field.

Explore the Simulator

SkyEye Monitor

Intelligent multi-tenant CCTV AI platform

8+ AI detection modules, unlimited cameras per site, 24/7 monitoring, 100% cloud or on-premise. Scales from 4 to 400+ cameras.

Explore SkyEye

Shatakshi Shipping Prototype

An indigenous counter-UAS detection system combining EO/IR sensor fusion, stereo-vision or LRF ranging, and live incoming direction or geo-coordinate tracking — built and field-tested with C3iHUB, IIT Kanpur, and Army units.

8-10
Fixed EO/IR Cameras
1-1.5km
Day Detection (EO)
400-500m
Night Detection (IR)
<200ms
Camera-to-Alert Latency
15 Hz
Direction Vector Output
<₹50L
Entry Price per Unit
Shatakshi EO/IR dual-camera prototype on rotating mount
Shatakshi dual EO/IR prototype — in-house build, undergoing calibration

two configurations including Stereo-Vision Detection on rotating platform with Live Geo-Coordinates and camera array integrated with PTZ camera and Kill platform

first configuration consists of The dual-camera stereo-vision rig triangulates detected aerial objects in real time, outputting live distance and GPS coordinates — not just a bounding box. Every detection streams a direction vector at 15 Hz to drive pan-tilt platform tracking automatically.

The second config ships with a camera array on a fixed platform with long range detection and PTZ camera for locking on to a target, this system integrates with a directed Kill system on a pan-tilt platform

This is the same sensor logic validated in the field: the screenshot below is an actual detection run, not a simulation.

Field-Tested, Not Just Lab-Tested

Shatakshi has been built and field-tested with C3iHUB, IIT Kanpur, and Army units, real outdoor conditions, real Army personnel, real validation. Patent filing under process on the core detection architecture.

Patent filed
Army field trials
Shatakshi field testing with Army personnel
Field deployment and testing with Army personnel on site

Two stages. One roadmap.

Stage 1 · Shipping Prototype

EO/IR Mid-Range System

  • 1.5 km sensor range, 3-4 km perimeter protection with multi system deployment
  • Stereo-vision / LRF ranging with live geo-coordinates
  • Patent on core technology filed
  • Field-tested with C3iHUB, IIT Kanpur and Army units
Stage 2 · In Development

Extended-Range RF + Vision Fusion

  • Long-range EO/IRST optics + 3D RF direction, 5 km detection range
  • Track-to-track correlation and multi-sensor data fusion (MSDF)
  • CNN/DNN automated target recognition and threat assessment
  • Low-cost interceptor drone solutions for the kill layer

Closing the loop: Detect, Track, Neutralise.

A Innovative interceptor drone concept, cued directly from Shatakshi's triangulated track — closing the loop from detection to neutralisation in one Indian-built system.

Capability

  • Cued from Shatakshi's triangulated track — Interceptor drones are launched to kill
  • Effective engagement radius up to 500m from launch point
  • Swarm-capable — multiple interceptors coordinate against saturating attacks

Deployment & Economics

  • Additional kill layer over the directed PTZ platform RF-jamming or machine gun mount
  • Ships as an optional kill-chain module bundled with Stage-2 Shatakshi
  • Under ₹1 lakh per interceptor; reusable launch and recovery platform
  • Can carry a detonating payload with proximity fuse, or a short-range missile net
Fixed Camera Array8MP bullet, single-EO or dual-EO/IR sensor, IP67, PoE — 10 units
PTZ Camera4MP, 40× optical zoom, 1km thermal range — 1 unit
GPU Processing ServerNVIDIA 96 GB VRAM, AMD 128 GB RAM, NVMe SSD
SoftwareFull Shatakshi counter-UAS platform — EO+IR fusion, PTZ tracking, ZMQ output
Pan-Tilt PlatformMil-grade, machine-gun mountable, microcontroller-driven per detector feedback
AMC12-15% of system value annually
Validated ByC3iHUB IIT Kanpur · Army units

Physics-true VR training —
the only trainer with detection in the loop.

Global VR trainers use gamified proxies. SkyAI trains operators on the exact same sensor logic they will field — real RF, acoustic, and optical sensor models, continuously calibrated with data from deployed Shatakshi units.

Immersive VR Simulation

Photorealistic airports, urban, and critical-infrastructure environments on Unreal Engine 5. Meta and Varjo headsets. Multi-operator network mode.

Detection Engine

RF, acoustic, and optical sensor models mirroring real hardware — continuously calibrated with data from deployed Shatakshi units.

Physics Engine

Aerodynamics, propwash, battery drain, wind shear, thermals, RF interference, and GPS spoofing modelled to military-grade fidelity.

AI Threat Scenarios

Adaptive swarms of up to 50 drones, stealth profiles, and electronic warfare. No-code authoring with versioned scenario libraries.

Evaluation Stack

Offline Monte Carlo validation of detection models, plus a live KPI engine: time-to-detect, false positives, miss rate, session scoring.

Hardware-in-the-Loop

API integration with RF and EO/IR scanners for hybrid live-VR exercises. Cloud-deployable and air-gap ready for classified sites.

One platform. Complete readiness. Zero live-drone risk.

SkyEye Monitor

An enterprise-grade, multi-tenant AI CCTV platform that deploys intelligent computer vision models on your existing camera infrastructure. Detect threats, enforce perimeter security, ensure safety compliance, and generate actionable insights — all in real time.

8+
AI Detection Models
∞
Cameras per Site
24/7
Real-Time Monitoring
100%
Cloud or On-Premise
4-400+
Camera Scale Range
3
Role-Based Access Tiers
SkyEye weapon detection AI model output
Live weapon detection — person + weapon dual bounding-box classification
SkyEye vehicle ANPR classification
Vehicle & ANPR module — type classification and number-plate read against watchlist

Eight Production-Grade Intelligence Modules

Each module is independently assignable per camera. Tenant admins dynamically configure which models run on each camera — no vendor lock-in, no wasted compute.

Civilian & Personnel

Detect and classify civilians and army personnel.

Person Detection

Multi-object tracking, occupancy counting, intrusion flagging.

Vehicle & ANPR

Detect vehicles, classify types, read plates against watchlists; classify army vs civil vehicles.

Face Recognition

Identify known individuals, flag unknowns, log all entry/exit events.

Zone Restriction

Mark safe/restricted zones per camera for persons or vehicles — useful in perimeter security.

Off-Hours Intrusion

Burglar alarm logic — any movement during off-hours triggers alerts.

Missing Guard

Alerts when a guard or worker is absent from post beyond the allowed time.

Alert Logging

Every violation logged to the database. Generate compliance reports on demand.

Smart Buzzer

Assign buzzers per camera — physical alert on AI-triggered violations.

SkyEye perimeter intrusion detection
Zone restriction & perimeter detection — multi-person tracking along a fence line

Enterprise Multi-Tenant Architecture

Built ground-up for multi-site enterprise deployments. Each tenant operates in a fully isolated environment with their own users, cameras, models, and data.

Super Admin

Creates tenants, sets camera & model quotas, manages platform-wide settings.

Tenant Admin

Adds cameras, assigns AI models per camera, manages users and alert rules.

Viewer

Views live feeds, browses reports, monitors alerts — read-only access.

Deployment Flexibility

  • On-premise, cloud, or hybrid deployment
  • RTSP/ONVIF compatible with existing cameras
  • GPU-accelerated inference engine
  • Scales from 4 to 400+ cameras

Smart Alerting

  • Buzzer assignment per camera
  • Push, email & SMS notifications
  • Configurable cooldown to prevent flooding
  • Evidence capture — frame + 10s video clip

Security & Compliance

  • Role-based access control (RBAC)
  • Full audit trail and alert history
  • Encrypted data storage & transmission
  • GDPR-ready tenant data isolation

Analytics & Reporting

  • Per-tenant compliance dashboards
  • Violation trend analysis over time
  • Exportable PDF/Excel reports
  • TimescaleDB for time-series analytics

Beyond our products,
we build what you need.

When your requirement doesn't fit Shatakshi, the Simulator, or SkyEye, our AI engineering team builds bespoke solutions — with the same sovereign-first, production-grade engineering.

Computer Vision

Custom computer vision beyond our flagship products — object detection, tracking, scene understanding, specialised classifiers.

  • Custom object detection and classification
  • Multi-camera cross-tracking and re-identification
  • Behaviour and activity recognition
  • On-prem deployment with TensorRT optimisation

AI Model Generation

End-to-end custom AI model development — from data collection and annotation through training, optimisation, and deployment.

  • Custom model training on your proprietary data
  • Data annotation and augmentation pipelines
  • Edge deployment on constrained hardware
  • Model versioning, monitoring, and refinement

Generative AI

Enterprise generative AI — document intelligence, conversational agents, content generation, domain-specific LLM applications.

  • Document intelligence and information extraction
  • Secure, on-prem conversational AI agents
  • Fine-tuned LLMs for defence and enterprise use
  • Retrieval-augmented generation (RAG) systems

Aerospace, AI, and Army
in one founding table.

SkyAI Technologies is an Indian deep-technology company building an indigenous airspace and perimeter security platform — sensor-fusion detection hardware, kinetic interceptors, physics-true VR simulation, and AI-powered CCTV monitoring. Developed in collaboration with C3iHUB at IIT Kanpur, a national Technology Innovation Hub.

Sarvagy Shukl

CEO / CTO

Ex-defence technologist; 16 years in aerospace, software, and simulation. B.Tech-M.Tech Aerospace, IIT Kanpur.

Dr. Anirban Roy

Chief of Tech & Hardware

Professor, Wadhwani School of AI, IIT Kanpur. 25 years in the aerospace industry.

Dr. Indranil Saha

Chief of Software & AI

Professor, Computer Science, IIT Kanpur. Formal methods and autonomous systems.

Maj. Virat Mishra

Chief of Strategy

15 years of experience across bureaucracy and the Armed Forces.

Nainesh Joshi

Chief of Business Dev

Defence and aerospace network spanning 7 countries.

Shefali Bajpai

Chief of Finance

15 years in finance across corporate and startup environments.

Mentors and advisors: Prof. A. K. Ghosh (ex-Aerospace Engineering, IIT Kanpur), aerospace and defence industry experts.

Recognised for indigenous innovation.

IIT Kanpur Unveils Anti-Drone System 'Shatakshi'

Developed by a team of researchers and engineers, Shatakshi integrates multiple layers of detection and countermeasures, including signal jamming, to detect rogue drones in real time.

MoD In-Principle Approval for National Military Drone Technology Hub

The Army Design Bureau has designated IIT Kanpur as the single point of contact for a ₹500 Cr National Military Drone Technology Hub, offering full-stack capability from design to certification.

IIT Kanpur Develops Indigenous AI-Based Anti-Drone System

The system was developed by Sarvagy Shukla, a former IIT Kanpur student who is now the CEO of SkyAI, addressing limitations of radar-dependent detection.

सीमा में घुसने से पहले खत्म होगा ड्रोन — IIT कानपुर का AI सुरक्षा कवच

News18 Uttar Pradesh coverage of Shatakshi's indigenous AI-based drone detection and neutralisation capability for border security.

Who We Serve

  • Armed forces and MHA paramilitary (base and border protection)
  • Airports — mandated drone-threat readiness
  • Critical infrastructure — refineries, power plants, ports, data centres
  • Police academies and training commands
  • State police and intelligence agencies
  • Government & judicial installations

Let's talk about your deployment.

Whether you're planning a counter-UAS installation, an AI surveillance upgrade, or a custom AI engineering project — we respond within one business day.

General Inquiries

For commercial, enterprise, and research collaboration inquiries.

Ready to see SkyAI in action?

Request a technical walkthrough, a live demonstration, or an on-site briefing at your location.

Request a Demo