AIOps Evolution Podcast | Season 2
Interview with Bryan Dell and Mohan Kompella from BigPanda
BigPanda has been a leader in the AIOps market for the last ten years. From day one the company has been built to solve problems. It’s evident that they are solving more than problems, but innovating the AIOps industry. They recently became the 800 pound panda in the AIOps market valuing over USD 1.4 billion. Additionally, BigPanda just announced they raised $190 million in funding to help them secure their position in the marketplace.
In this market where everyone is pedaling AIOps, BigPanda knows they’ve got to stand out.
So, what brings customers to the AIOps table and what are the differentiators they’re looking for? In this episode of the AIOps Evolution Podcast, Sean got the chance to discuss this question with BigPanda’s former Chief Financial Officer, Bryan Dell, and current Product Marketing Lead, Mohan Kompella.
The good and the bad of investing in AIOps
There’s no question that AIOps has gained traction within the last five years, especially driven by the pandemic and remote work conditions. Bryan said, “The pandemic really shed light on the value that ITOps brings to an organization, particularly as everybody had to figure out how to work remotely and were scrambling to make all their systems operate effectively.” The result of that is that ITOps has earned a much more strategic seat at the table in the executive boardroom. And they have the power of AIOps to help them be more effective.
That’s the good news, according to Bryan. The bad news is because AIOps is growing so popular and valuable to organizations, it’s also become a very broad, bloated, confusing category for consumers. Bryan states, “So every vendor in the monitoring space, in the ITSM space, in the ticketing space, in the collaboration space, in the on-call space, everyone is saying they do AIOps; and everybody is telling enterprises and customers that their version of AIOps is what they need right now.” This is extremely overwhelming for IT Operations leaders who are looking to make strategic investments in AIOps. Many are left asking, “How do I make sense from the noise?”
Differentiators Driving AIOps Investment
Since there is no shortage of companies that market AIOps, we asked Mohan and Bryan what sets an AIOps company apart from the pack?
Ease of AIOps Implementation
If there is a feature to look for in your AIOps product, it is the ease of the implementation process. Mohan related that enterprise users looking to take advantage of AIOps value ease and out-of-box learning. This capability employs logic that the ML uses to correlate errors and make it meaningful to humans. Users test and preview results before deploying that in production. Enterprise users develop trust with a system where they can see what AIOps is doing, test it, and edit it.
Customers value transparency, testability and controllability
When it comes to machine learning, engineers want to know what it’s doing and how it works. Mohan said, “AIOps customers are looking for a product that provides full transparency, testability, and controllability.”
It starts with the data collection. The console collects the data and users review that information to view the correlation logic (pattern) in plain English. Users can select a given time period to test that data against. After a quick analysis, the console is able to return a hypothetical outcome of what would have happened to that past data set had they been using the data correlation logic. It’s a “what if” machine. Users go back in time and replay that data, applying a new correlation to see what could have been. It is not simulated data, but data you’ve collected.
Mohan notes that customers looking at investing in AIOps want to stay in the driver’s seat. That should be at the core of the system. Users don’t have to build their own ML model. Instead, they deploy AIOps into their environment where pre-trained models hit the ground running and start reducing noise in about 6-8 weeks. He says that over the course of those first few weeks, there is typically a 70-75 percent noise reduction rate as the AI learns the environment.
New tools added are not a problem either. AIOps software begins to automatically parse the data periodically to suggest new correlation patterns. Mohan reiterates that ‘the keyword is suggest.’ An AI should not make autonomous decisions for users. It should suggest a new correlation pattern and then the user can examine it, test it, make changes, and deploy it in production as they see fit.
Choose Your Own AIOps Adventure
Users want to choose the patterns and what happens with the data. They don’t want to wait for the machine to update the new pattern. Instead they can play with the system and update the UI with a new pattern as they see fit. In five to ten minutes (what would have taken two weeks), users can create a brand new correlation pattern. That really sells folks on AIOps where they feel like they are not dependent solely on the machine process. They are using the machine to help them do their job better and increase productivity.
Neutral agnostic vendor approach
A typical enterprise has about eight to ten monitoring tools as well as five to six sources of topology and change data. That means that most enterprises looking to invest in AIOps will be looking for a tool or a platform that can integrate with multiple tools (20-25 generally).
With an event enrichment engine, it adds operational and topological context at scale as well as process millions of alerts everyday. In real-time, it enriches the incoming alerts stream and unifies fragmented tools across several environments. This fulfills what many customers are looking for–something that can work effectively in hybrid environments.
Investing in AIOps, the IT game changer
The benefits of AIOps are endless, but it all depends on how IT teams apply it. As Bryan and Mohan said, not all AIOps tools are created equal. That’s why it’s critical to curate AIOps products to strategically place control in the hands of the users, not the other way around. That’s the real differentiator for bringing customers to the AIOps table.