What is Anomaly Detection?

What is ‘anomaly detection’?

It’s 5:49AM, and you’ve got a busy day ahead of you. You didn’t sleep properly, and you have 3 hours till that big presentation you need to prepare for. You open your kitchen cabinet, only to find that you’re out of coffee grounds, naturally. Quickly, you pull out your laptop and go onto your regular online grocery delivery service and select your favorite coffee grounds.

But when you click Buy Now, you find that the payment is taking forever. You get a message on your phone saying the money for the coffee has been deducted, but the website doesn’t give you a confirmation on your order. Tired and irritated, you start to sleepily prepare the presentation anyway, but the sleep deprivation is apparent throughout the rest of your day.

This unfortunate incident, technically called an anomaly, could have been avoided if the banks at both ends were using AI driven monitoring systems!


In reality, the simple Buy Now button that you click actually abstracts a web of highly complex interactions, distributed across various systems within the banks, payment gateways, etc. To make even a single purchase of a coffee happen smoothly, you need to be able to monitor the multitude of systems in a holistic and end-to-end manner, because when something goes wrong it can be highly difficult to identify where the problem is!

Companies have to monitor and process millions of data points to maintain these transactions, a task which is pretty much impossible for humans to do manually. At VuNet Systems, we monitor over 2.5 billion transactions a month– connecting with some of the largest banks and payment gateways to ensure that every transaction is smooth- a task even more important in a Covid-19 world.

In order to monitor and process all this information, companies need to use monitoring systems that automatically monitor all of the systems in an enterprise and give them updates on how each system is functioning so that they can quickly detect anomalies and proactively take corrective action. When things are running smoothly, you as a customer know it (your daily dose of caffeine is fulfilled!), and the monitoring system can see this as well. But when you run into these seemingly random issues, usually you can’t pinpoint the exact problem. 

Automated AI driven monitoring systems organise all this messy and voluminous data into a smaller tractable set of metrics using intelligent algorithms and can also let companies know when something is wrong; just like a rather annoying morning when you’re unable to pay for much needed caffeine. Or, remember when you tried to add those jeans to the cart and it just wouldn’t get added? Usually when this happens, one of the processing systems or associated infrastructure has some sort of issue, and an AI driven monitoring system is able to see that and alert the company. 

Being the only Indian company that provides a unified AI driven Monitoring System to several major banks and financial institutions, VuNet Systems detects this pesky problem with your payment and alerts the bank immediately- before you know it the money has been returned to your account, and your day seems to go a lot more smoothly. Any digital business, be it an e-commerce platform or a bank, not only manages a lot of data, but also needs to perform a lot of analytics to make their business run without problems. And for the last few years, VuNet has been doing exactly this. The end-to-end integrated AI based platform developed by VuNet connects user experience, customer journeys, business transactions, infrastructure and IT operations. This allows a comprehensive bird’s eye view of the IT operations to derive real time business and operational insights on the platform. 

This is a much more difficult task than you might think! When you have transaction systems of high volume, you need alerts that are proactive (so they tell you something is wrong immediately), predictive (alerts that can pinpoint what is likely to go wrong in the future), and diagnostics (something that analyses the root cause of your problems). At VuNet we do all three using state-of-the-art AI algorithms.

Hope you now understand why anomaly detection is such an exciting and important field of work which impacts your day-to-day life in more ways than you can imagine!

With this understanding you will be able to appreciate a further deep dive into Anomaly Detection concepts and a view into the complex adventures VuNet goes through daily, which we will bring to you in this blog series on Anomaly Detection.

In the next blog, we will take you through the basic concepts of anomaly detection and also some guiding principles that we follow at VuNet Systems.

VuNet Systems is a deep tech AIOPs startup revolutionizing digital transactions. VuNet’s platform vuSmartMaps™, is a next generation full stack deep observability product built using big data and ML models in innovative ways for monitoring and analytics of business journeys to provide superior customer experience. Monitoring more than 3 billion transactions per month, VuNet’s platform is improving digital payment experience and accelerating digital transformation initiatives across BFSI, FinTechs, Payment Gateways and other verticals.