Augtera also announces Zero Day Anomalies for syslog, detecting unusual patterns when they first occur.
PALO ALTO, Calif., April 27, 2022 /PRNewswire/ — Augtera Networks, the industry leader in AI/ML-powered Network Operations platforms, today announced a major innovation in streaming log analysis, utilizing real-time natural language processing (NLP). The next evolution in log analysis is moving from text matching to understanding meaning, both the semantics and the context of text, in real-time, for proactive insights. Text matching is not scaling. It is brittle, hard to maintain, and generates more noise than operations teams can meaningfully digest. Natural language processing, combined with other AI/ML techniques, enables innovative new capabilities while also reducing noise.
“AI/ML is proving to be a major step forward in network operations, yet many Enterprises do not have budget and resources to successfully develop and maintain in-house implementations that are performant, scalable, and extensible,” said Nick Lippis, ONUG Co-Founder and Co-Chairman. “Commercial solutions from companies like Augtera Networks are filling the gap. Automated detection of new log messages and patterns like Augtera’s Zero Day Anomaly capability is a great example of the real-time anomaly detection that Enterprises will increasingly need across multiple network vendors and domains. This is particularly important as Networking and Security become more entangled.”
The first feature that Augtera is announcing based on NLP is Zero Day Anomalies for syslog, which also leverages clustering and machine learning. The conundrum of the current log processing paradigm is that an anomaly must be already well known before a text-based filter can be created to detect it. Even then, text-based filters based on approaches such as REGEX can be noisy and hard to maintain in the face of constant change and variation across multiple vendors.
“This is a new era in log analysis” said Rahul Aggarwal, Founder & CEO, Augtera Networks. “Operators increasingly want to move away from reactive historical mining of log data to proactive real-time insights. Our ability to discover meaning from log data and utilize it for real-time Zero Day ML Anomalies catalyzes this shift. This gives operators the peace of mind that an unknown hardware or software issue can be detected automatically before it impacts the rapidly evolving network infrastructure that creates even more unknown unknowns”.
These new additions to Augtera’s Network AI platform are the result of over a year of R&D and are powered by new proprietary algorithms. These algorithms build ML models that leverage NLP over Big Data. They are already deployed in production to predict Zero Day Anomalies in environments with hundreds of millions of syslog messages an hour.
“The ability to understand log message semantics will be the foundation of many future capabilities,” said Bhupesh Kothari, Co-Founder & VP of Engineering for Augtera Networks. “Augtera has leveraged the team’s experience in embedded software development, to write from scratch, high-performance and cost-effective AI/ML algorithms, that scale to billions of streaming log messages per day, without message loss. Engineering challenges for real-time streaming AI/ML are significantly different than those for historical data lake mining. Our implementation supports the largest data centers and networks.”
Augtera Network’s algorithms and models are optimized, based on real-world deployments, for network operations. While the Augtera platform supports the most demanding and largest network operations environments, it has also been developed to provide immediate value upon initial installation, without configuration. Our focus is network operations automation through automated anomaly detection, incident prevention, and trouble-ticketing.
Zero Day Anomaly Detection for Syslog is immediately available as a base capability in the platform, at no additional cost.
About Augtera Networks
Augtera Networks stops the noise, enables proactive operations, and prevents incidents, for Enterprise and Service Provider networks. The first AI/ML-powered network operations platform, Augtera is being used by hyperscale cloud platforms, financial institutions, communications service providers, managed service providers, and enterprises in multiple verticals. Additional information can be found at www.augtera.com.
SOURCE Augtera Networks