Network AI uses purpose-built AI/ML to eliminate alert fatigue, reduce incidents by 90%+, and detect future incidents. Network operations teams see more, while doing less. The combination is a virtuous circle of ongoing improvement: a dramatic reduction in the number of incidents and a dramatic reduction in the time to respond to incidents.
Network AI is multi-layer, from physical layer to TCP, cloud logs, and beyond. Networks are different than compute, storage, and applications. They require purpose-built algorithms because patterns, constructs, and behaviors are different. Even networks are different from each other, as are interfaces. Augtera Network AI is built by network experts for network experts.
Network AI leverages both community-based classifiers for experience-based high-fidelity signals, as well as learned full and gray failures, detected by 9+ Augtera purpose-built algorithms. Anomalies that have already impacted KPIs are responded to quickly, and anomalies that are likely to cause future failures can be prevented. Some of the many use cases include optical signal degradation, VLAN BPDU errors, IP traffic anomalies, BGP session errors, Impact of fabric congestion on applications, latency packet loss anomaly detection, and more.
Learn how Augtera Network AI reinvents network operations by applying machine learning to mine data from various parts of the infrastructure to produce AI insights to be consumed either by humans or automation tools.
Presented by Rahul Aggarwal and John Heinz at Networking Field Day, May 6, 2022.
Built to ingest and normalize every data point via SNMP, syslog, telemetry streaming (OpenConfig based gRPC/gNMI or protobuf), sFlow, IPFIX, ERSPAN, REST APIs, TWAMP, Augtera and 3rd party synthetic probes, Augtera streaming API, Kafka and custom data sources. Your custom JSON data can be dynamically normalized.
All nodes, relationships and attributes of the network are automatically discovered across physical and virtual topologies including data center, multi-cloud, SD-WAN and WAN network domains. Protocols and network technologies that are supported include Layer 2, IGP, BGP, MPLS, L3VPN and EVPN. A very wide range of network constructs and components are supported.
9+ proprietary, network-specific, ML algorithms built from the ground up that autonomously learn the patterns of all network infrastructure, application flow and SLA timeseries metrics as well as syslog text, build online models and predict operational anomalies in real-time. This enables proactive detection before brewing network issues and grey failures explode into downtime and business impact.
Example – Augtera learns the normal seasonal pattern with cyclical spikes and predicts the unexpected metric behavior as an anomaly
Proprietary machine learning algorithm that is multi-layer network topology aware and correlates network events and Augtera anomalies across multiple data sources automatically. This enables operators to proactively receive notifications of correlated network issues with high fidelity context, further reducing ticketing noise and enabling rapid root cause analysis and remediation.
Example – Auto-correlated Incident with 11 synthetic probe RTT / Loss and interface packet drop anomalies in the data center fabric
The Augtera Networks platform automatically classifies log data and quickly detects pattern changes and behavioral anomalies that other systems can’t find. Augtera also leverages proprietary NLP and Machine Learning to find zero-day anomalies in log data. These anomalies enable automatic detection and notification of operationally relevant logs when they first occur.
Read more about Augtera syslog analysis.
Physical and virtual topology visualizations with an overlay of network events and anomalies across multiple data sources with time machine support. The Augtera Networks platform assists the operator to rapidly diagnose the root cause when there are complaints from application.
Simplified and flexible workflows for different teams driven by operator intent supporting ad-hoc analytics, notifications (slack, syslog, kafka), and automated ticketing integration (Service Now). Out of the box and custom Views with rich metadata aware filters (e.g., choose to notifiy only certain types of anomalies on certain types of dervices) are used to define what types of anomalies, events and auto-correlated needles consoles, ticketing systems, and automation systems should be notified about.
Real-time ad-hoc visualizations of topology, metric, event and syslog analytics, anomalies, auto-correlated incidents.
The destination is automation. This requires a platform that is API-native in every direction to enable devOps workflows and take the last step from proactively detecting an issue to taking action.
Designed with support for operational capabilities and scale that enterprises and providers need.
Noiseless anomaly detection, root identification, and policy-based notification all combine to deliver operationally relevant needles, continuously, and proactively, from a data haystack that was previously only mined manually, infrequently, and reactively.
eCommerce hyperscalers, service providers, leading financials, retailers, and other Enterprises are adopting the Augtera Network AI platform for its ability to transform network operations. Detection, mitigation, and remediation times are decreased by 60% to 90%, with network engineer action occurring in less than one minute.
Billions of data points per day are analyzed in real-time, continuously, and automatically. Noise is eliminated, skilled operations resources are focused sooner on fewer, highly relevant incidents. Operations teams can also have more time, to proactively respond to gray failures and other emerging trends, preventing future incidents.
A new platform, designed from the ground up, to streamline the process of getting to mitigation as fast as possible.