Topology discovery
Build a live network map from device and routing information.
Smart India Hackathon / Network systems
NetLine
A network discovery and monitoring system built for GAIL India, winner of Smart India Hackathon.
The operator view
Make a changing industrial network visible before it becomes an incident.
Industrial network teams need to understand topology, device state, and traffic changes without manually reconciling equipment and interface data.
NetLine continuously discovers the network and turns low-level protocol information into a live topology that operators can inspect, monitor, and use during investigation.
Change monitoring
Track device, interface, and adjacency changes as the network evolves.
Traffic visibility
Surface flow and operational events alongside topology context.
Operator workspace
Explore the network through an interactive D3 interface with live updates.
Build notes
The graph was only the last step.
I built across network discovery, backend services, data modeling, real-time delivery, and the interactive topology experience used for the final system.
- Our team built NetLine around the question operators actually ask during an incident: what changed in the network, and where?
- I worked across the discovery data, backend services, topology model, WebSocket updates, and the D3 interface used to inspect the network.
- The build connected protocol-level information to an operator-facing map instead of leaving SNMP, routing, and flow data in separate views.
Treat topology as changing state
The map is not a static diagram. Devices, interfaces, and adjacencies have identities and updates so a change can be attached to the object it affected.
Normalize before drawing
Protocol collectors return different shapes. The backend turns those records into one topology model before D3 receives them.
Stream relevant changes
WebSocket updates keep the visible network current without replacing the entire graph every time a signal arrives.
Operational data becomes useful when every event is attached to the topology it changed.
Limits of the hackathon build
The public repository preserves the hackathon build, and its README was never rewritten into proper system documentation.
The demo could not reproduce every vendor, permission model, and failure mode found inside a live industrial network.
Production SCADA access would require much stricter credential handling, network isolation, audit controls, and deployment validation.
TypeScriptD3FastAPIPostgreSQLSNMPv3WebSockets