DriveNets builds cloud-native networking software that handles large-scale routing for service providers and AI infrastructure operators. The company's core technology disaggregates network functions from proprietary hardware, running instead on white-box infrastructure with open interfaces - essentially applying cloud architecture principles to carrier-grade networking. By the company's metrics, its software now supports over 30% of US internet traffic through deployments at AT&T, Comcast, and other major providers. The technical domains span scalable network infrastructure, disaggregated routing protocols, and the operational economics of running internet-scale networks without vendor lock-in.
The architecture matters because traditional routing platforms bundle software, silicon, and chassis into expensive, rigid systems that scale vertically. DriveNets' approach separates control plane software from commodity switching hardware, allowing operators to scale horizontally and update components independently - a shift that changes both capital expenditure patterns and operational flexibility for networks moving petabytes daily. The company positions this as infrastructure modernization comparable to what happened in compute and storage a decade ago, now applied to the routing layer where resilience, throughput, and cost per bit determine competitive positioning for providers.
Headquartered in Israel with distributed teams globally, DriveNets operates in the telecommunications and cloud infrastructure verticals where network reliability directly impacts service delivery at scale. The company describes itself as entrepreneur-led, with a technical culture emphasizing practical engineering over incremental improvements - teams work across time zones on problems where milliseconds of latency or percentage points of packet loss have material business consequences. For engineers, the environment involves dealing with the complexity of production networks carrying real user traffic, where design decisions around protocol implementation, state management, and failure modes play out at internet scale rather than in lab conditions.