Observability: The Secret Hero At The Back Of Every App

Observability: The Secret Hero At The Back Of Every App

The volume of users and speed at which operations run in India make it imperative for platforms to provide exceptional digital experiences. Ashan Willy, CEO of New Relic, tells us how their observability platform is the backbone of apps we use in our everyday lives.

Tsunami CostabirUpdated: Monday, May 13, 2024, 10:39 AM IST
article-image

In March 2022, New Relic, a US-based observability company, set up operations in India with an office in Bangalore. Since then, the company valued at USD 6 billion has heavily invested in the country, setting up an Innovation Centre and a new office in Hyderabad. 

Observability provides visibility into how systems and apps operate, allowing engineers to monitor performance, understand user behaviour, diagnose issues and optimise functionality effectively. 

Used in popular apps like HealthifyMe, Swiggy, BigBasket and on financial and OTT platforms, New Relic’s software touches almost every aspect of our lives. With smartphone adoption in India reaching close to 71% in 2023, the elimination of disruptions in digital experiences has never been more important. 

What is the India potential you’re looking to tap into?

India is a growing investment space given the talent. The India market, too, is our fastest-growing market where we have over 12,000 customers. We service both digital-native and traditional businesses and since India has been at the forefront of everything digital, it gives us a bit of an opportunity here. Lastly, a lot of observability operators for large MNCs in the US sit in India. So if we service this market well, it will help us globally.

What are some key areas of difference between India and global markets?

Models in India are different. Like quick commerce for example, having services and deliveries at your doorstep in nine minutes is something that doesn’t happen in other parts of the world. And when you look at the volume of users and user concurrence, it becomes critical for companies to ensure that it doesn’t affect the platform functioning. So the speed and scale of operations necessitate that applications are really robust. 

What has led to the growing importance of observability?

New Relic was set up in 2008 by Lew Cirne, a pioneer in the development of Application Performance Management. The way applications were being written was getting more complex and he realsied that observability can be done from the cloud. For the various services and microservices they offer, an end-to-end monitoring of the application is really important. And today, with AI being embedded into applications, different performance capability characteristics need to be observed.

What does the future of observability with data security look like?

For effective observability, we do need data and have a responsibility to make the data secure. The cyber security industry is working to stop vulnerability and attacks at an early stage rather than waiting to find them later downstream. Interestingly, observability has a unique opportunity to disrupt that security market. We conduct vulnerability management and penetration testing while developers are building new software. So when the applications are already tested and secured, the role of the security industry will change. We’re trying to move the security testing to a step ahead of where it is. 

How much of the data that gets collected gets turned into actual insights?

We store about 1 exabyte and transact 2.5 exabyte of data per day. To put that into context, it’s 122 million Bollywood movies worth of data which would take you 42,000 years to watch. We process massive amounts of data to give insights. But yes, we’re moving into an observabilty 3.0 phase where we’re looking at being intelligent about the amount of data being collected. That way, we can work on getting the same insights with lesser data.

What are some observability solutions in AI that New Relic is working on?

With AI, hallucinations and biases are an issue. The biases can be caused by either model being used or the data we’re feeding it. As observability vendors, first we can monitor the integrity of the data that the large language model (LLM) is consuming. And second, LLMs are building into the model their own signals of biases. So once the model exposes that bias, we can pick up on it.

RECENT STORIES

Breaking Into Women's Minds With Pee Safe

Breaking Into Women's Minds With Pee Safe

Do I Still Need An Antivirus?

Do I Still Need An Antivirus?

Going Cookieless Could Be A Good Thing

Going Cookieless Could Be A Good Thing

How Amateur Sports Leagues Can Turn Profit-Making

How Amateur Sports Leagues Can Turn Profit-Making

Zoreko Is Pushing VR In Arcades

Zoreko Is Pushing VR In Arcades