Last updated 06-28-2026
Category:
Reviews:
Join thousands of AI enthusiasts in the World of AI!
WAN-AI
WAN-AI builds the Intelligent Packet Scheduler (IPS), an autonomous packet scheduler aimed at enterprise 5G networks. It targets operators and network teams who need guaranteed throughput and latency on individual data flows, not averaged best-effort service across the whole network.
The product pairs a physics-accurate simulation engine with machine learning in what WAN-AI calls Physics-AI. A closed loop senses live traffic, forecasts demand, computes optimal scheduling parameters, and applies them every ten seconds without manual config changes. The scheduler runs on a PIFO packet scheduler at line rate on commodity hardware.
WAN-AI is built for telecom operators, enterprise 5G deployments, and research platforms that need Level 5 network autonomy. In June 2026 the company began deploying IPS on the 6G-SANDBOX Berlin platform.
Enforces minimum throughput and max latency per flow, per traffic class
Closed loop runs every ten seconds with no human config changes
PIFO scheduler at line rate on commodity hardware
flow_monitor samples live traffic every five seconds
net_modeler simulates 250x faster than conventional network tools
params_dnn infers optimal scheduler settings in 5 milliseconds
Guarantees per-flow throughput and latency targets instead of network-wide averages.
Fully autonomous closed loop re-optimizes scheduling every ten seconds without engineer intervention.
Runs at line rate on commodity hardware via a programmable PIFO packet scheduler.
Physics-accurate simulation engine reports 250x faster performance than conventional tools.
No public pricing; enterprise access requires scheduling a demo.
Focused on enterprise 5G packet scheduling, not general network management.
Limited public documentation beyond the single marketing site.
What is WAN-AI IPS?
WAN-AI IPS is the Intelligent Packet Scheduler from WAN-AI. It is an autonomous packet scheduler for enterprise 5G that enforces per-flow SLA targets on throughput and latency in real time, using a closed Sense-Predict-Optimise-Act loop.
How does WAN-AI optimize network traffic?
WAN-AI IPS runs four stages continuously: Sense captures live flow data every five seconds, Predict forecasts connection rates, Optimise uses a physics-accurate simulation engine and a trained params_dnn model to compute scheduler settings in 5 milliseconds, and Act applies cap, burst, PIE queue management, and weighted fair queuing parameters at line rate.
Does WAN-AI offer a free plan?
WAN-AI does not publish pricing on its website. Access is through a demo request form where you submit contact details to schedule a product demo with a WAN-AI consultant.
Who is WAN-AI built for?
WAN-AI targets enterprise 5G networks and operators that need autonomous packet scheduling with per-flow SLA enforcement. The company is deploying IPS on research platforms such as the 6G-SANDBOX Berlin environment.
Where is WAN-AI based?
WAN-AI is incorporated in both the United Kingdom and the United States, as stated in the site footer. The company lists UK, US, and EU presence on its homepage.
What scheduling parameters does WAN-AI IPS control?
WAN-AI IPS applies cap, burst, PIE active queue management, and weighted fair queuing parameters to every packet on every flow at line rate. The scheduler is programmable and adapts to different SLA structures and traffic profiles.