When we started Augtera we had a mission to leverage Machine Learning and deep analytics to transform the networking industry. Today we are launching our company and our industry-first Network AI platform – a major milestone towards that goal.
We are launching with several Fortune 500 and global customers, including Orange and leading e-commerce and financial enterprises, who have deployed Augtera in production at scale with significant quantifiable benefits. And with $18M of capital raised to date from top tier venture capital firms Bain Capital Ventures, Intel Capital, Dell Technologies Capital, and Acrew Capital.
We have also teamed up with Dell and Netone to integrate our platform with the solutions that they are bringing to the market around hybrid-cloud, disaggregation in the data center and software defined networking.
By bringing to market our platform, Augtera is reimagining network performance and management using our optimized AI algorithms to provide proactive visibility, prevent network incidents, and dramatically reduce resolution times in multi-cloud, data center, WAN and SD-WAN networks.
We started with an intuition that the networking industry had fallen behind in the adoption of Big Data and AI technologies and that an AI platform purpose-built for networks could transform network operations, orchestration and design.
As we started discussions with network engineers and operators our intuition was not only validated but we also realized that the use cases for such a platform were unlimited. It also became very clear early on that such a platform had to be “purpose built” for networks because it had to understand network semantics including rich and complex network topologies and diverse data types and distributions. Our team had deep expertise in networking, having pioneered and built industry changing MPLS and VPN technologies and products, and in addition we had spent several years working with analytics and Machine Learning. We had the right team, were prepared to undertake fundamental R&D and had the patience needed to gather large scale production data and then build our algorithms.
New Network Intelligence
We decided to focus our energy initially on network operations as entropy and fragmentation were increasing with the adoption of hybrid cloud and software defined networking. Fast forward to the present and this trend has only accelerated and network operators are facing increasing urgency to solve burning issues: lack of proactive visibility, overwhelming noise, silos between networks and applications.
Addressing these issues requires network intelligence that hasn’t existed before.
Pattern recognition is how a new intelligence takes shape. At Augtera by ingesting and analyzing every bit of data from every source we uncover patterns at scale. This leads to entirely new data and insights and hence creates a new network intelligence that is the foundation of our platform.
Problem and Solution
Network operations is very reactive today. Operators scramble to fix issues when customers complain and applications suffer. Augtera on the other hand applies anomaly detection algorithms purpose built for the network to automatically find what’s normal, to find problems when they are brewing, before they reach the boiling point and before things break. This allows operators to detect problems, grey failures and unknown unknowns proactively, thereby avoiding network incidents.
There is so much alert noise, lack of context and fragmentation today that when things break it takes significant time and manual effort to separate the signal from the noise, find the root cause and resolve issues. With Augtera the signal-to-noise ratio is much higher and operators can detect issues that escape prevention much faster. Augtera’s AI algorithms provide so much context and cross-layer correlation that operators can then find the root cause and the resolution for these issues far quicker. This is what we call AI Driven Resolution. We believe that as operators get comfortable with this, some of the resolution work flows can be automated, thereby reducing the mitigation times even further.
Augtera does this across the application and network layers allowing modern enterprises to break silos, reveal if the network is the cause of application issues and to get insights needed to tune the network for application performance.
Our network intelligence platform is already preventing 40-50% of service impacting network incidents for customers with 90% faster detection of critical issues and 50-60% reduction in resolution times. We believe that these numbers have significant room to improve with constant online learning and as the platform matures.
Our microservices based platform can be deployed either on-premise or in the cloud or in a hybrid mode. We are designed to handle all network architectures and are already deployed in production in hyper-scale data centers and private cloud, WAN, SD-WAN and hybrid-cloud. We are optimized for massive scale and the world of telemetry streaming. At the same time we have spent significant effort scaling SNMP and syslog data ingestion and Machine Learning pipelines and our AI algorithms are providing customers high value across these technologies that are still widely deployed across networks.
When we started Augtera we realized that network operations tools were stuck in the manual and static mindset of the past. To move the industry to the modern era we had to completely rethink this and build an intelligent system that learns automatically from the diverse network specific data distributions, understands network topology, the applications that run on that topology and provides high fidelity actionable results. This had to be done economically at scale and it was important to architect this as a platform that can integrate with the IT ecosystem and modern DevOps tools.
Our progress is the result of our team’s focus on solving high value difficult problems empowered by our investors and the trust placed in us by our customers. Thank you all. We look forward to the next chapter in our journey together.