Chakravyuh is a live geospatial command platform built for the Indian intelligence theater. It pulls open-source signals from social media, airspace transponders, maritime feeds, and news sources, correlates them through an NLP entity engine, and renders everything on a single tactical map. Operators get a 360° operational picture without switching tools or waiting for manual reports. The name comes from the ancient Indian battle formation that, once entered, could not be escaped. That is the standard we hold threat signals to.
Ingests and classifies signals from Twitter/X, Instagram, Telegram, Reddit, and live RSS news feeds. Each post is parsed for geolocation, named entities, and a risk score, then pinned directly on the tactical map. When signals cluster around a location or keyword in a short window, the platform flags it as a potential coordinated event for operator review.
Pulls live ADS-B transponder data every 10 seconds and plots real-time flight positions across the Indian subcontinent. Operators see callsign, country of origin, altitude, and velocity for every tracked aircraft. Unscheduled flights, low-altitude movements near sensitive zones, and border-crossing activity are immediately visible on the map.
Tracks vessels across the Arabian Sea, Bay of Bengal, and the wider Indian Ocean Region using AIS transponder feeds. Each contact displays vessel name, type, speed, and heading. Military contacts, unclassified vessels, and ships operating near India's Exclusive Economic Zone are priority-flagged for operator attention.
Every ingested document runs through a Named Entity Recognition pipeline that pulls out Persons, Organizations, and Locations. Those entities feed into a knowledge graph where edges represent co-occurrence relationships. Connections between actors, events, and places that would take an analyst days to piece together manually become visible in seconds.
Operators can query the platform in plain language to get fast answers across all active intelligence layers. Ask about a specific location, a vessel contact, or an entity in the graph and the system cross-references live feeds to return a focused briefing. It handles the data synthesis so analysts can focus on decisions, not search.
Every node on the map carries a risk score across three tiers: Critical, High, and Moderate. Animated tactical reticle markers draw the operator's eye to active threats first. During high-volume events such as civil unrest or influence campaigns, nearby signals are clustered so the map stays readable and actionable.
Chakravyuh runs on a multi-source ingestion pipeline that pulls from social media, aviation transponders, maritime broadcasts, and open news feeds, normalising everything into a single intelligence schema regardless of source format.
Each document passes through an NLP processing layer that identifies people, organisations, and locations, then encodes them as nodes in a knowledge graph. The edges of that graph represent co-occurrence relationships between entities, making it possible to spot connections between actors and events that would otherwise stay hidden.
The frontend renders this graph geospatially, placing entity nodes at or near their real-world coordinates on a dark tactical map. Aviation and maritime layers refresh on short polling intervals and render as animated markers so movement is immediately visible.
The platform is hosted on secure cloud infrastructure with global edge delivery, keeping response times fast regardless of where the operator is accessing from or how many sessions are active concurrently.