Rick Hamilton has been CEO for Infovista since September 2023, quickly making the step up to the job after being brought in as deputy CEO six months earlier. Hamilton’s previous job in telecoms was as GM of Blue Planet’s automation software business. After six years at Blue Planet, which was bought by Ciena in 2015, and also previous decent stints at both Cisco and Juniper, Hamilton thought he was “retiring from this corporate world”.
He planned a career change as a pilot.
“I got hired by Southwest and was ready to go. And then I got a call from our [Infovista] shareholders saying, ‘Hey, we’d like to move the company in a different direction. Are you interested?”
So what attracted him back from the skies to Infovista, a company that provides network automation, analytics and service assurance capabilties to network operators, giving it a certain amount of crossover with previous home Blue Planet?
Infovista is owned by Private Equity company Seven2, perhaps better known by its former name Apax Partners. It had pursued a strategy of buying different comms businesses in the test, assurance and planning software space. In 2012 it bought Mentum, and its network planning software. That was followed a year later by Aexio, an optimization software company. In 2016 it acquired TEMS, a well known network test business, from Ascom. In 2021 it added Empirix, bringing in its assurance capability.
The company had structured its products into three areas – RF planning, test, assurance – all of which could sit on a common cloud-native platform called the Network Lifecycle Automation Cloud Platform.
“Frankly,” Hamilton says, “that was the first bit, trying to get my head wrapped around how all these companies come together.
“That was kind of the start of it all. So the plan was to not unlike any PE backed firm – unashamedly to get ready for IPO – what’s the exit? Come in, take a look at the asset, the value that we think we’ve created, how do we propel the company forward for whatever the next chapter will be?”
We’re going to unify our product strategy under a more cohesive brand presence and take that latent value that we have and present it to the market in what will become pretty interesting ways.
Hamilton says that in his opinion, Infovista has a lot of under-exploited “latent value” that it can bring to the market.
“We’ve told our story up to this point – that you can plan, test, assure. Now when you think about the world of AI and how all those things can come together, we have unique advantages and skills, experiences and technologies we’ve developed over a long period of time. One plus one plus one doesn’t equal three – it should equal four of five.”
It is Infovista’s NLA platform that can be the multiplication engine for those capabilities – if the company itself can align as a businesses to deliver on those benefits..
Hamilton describes NLA as an architecture that “abstracts core capabilities” like analytics, automation and machine learning (ML) engines onto a common platform that then powers Infovista’s products.
“This is an architecture that allows us to interface in a standards-based way down into the infrastructure and up into the OSS. So that was all built as a part of NLA. But we hadn’t really monetised well because, frankly, the company thought of itself in those three silos (planning, test, assurance) and that’s the way we went to market.”
Despite that, Hamilton thinks that the drive towards AI-led automated network operations gives Infovista an opportunity to leverage its platform approach.
“We spent an enormous amount of money in what, in the AI context, are going to be very valuable assets. So we’re going to unify our product strategy under a more cohesive brand presence and take that latent value that we have and present it to the market in what will become pretty interesting ways.”
Automation and AI ambition
One of the key opportunities for companies like Infovista going forward will be to enable greater automation of network operations, driven by AI. That relies on access to data, and then the ability to structure it so that ML and AI engines can make use of it at the aplication level.
At TM Forum Infovista is talking part in a Catalyst project with BT, looking at very specific cases where automation is playing into the assurance stack – automation that comes out of an analysis of a problem.
“So we detect a problem, we analyse what the problem really is, and we can pass information to an Automation Engine that they control. It’s kind of rules-based but that, in a parochial way, is a good example.
“You can apply this stuff in a number of places – think customer care as an example. How do you automate the interaction you have with your customers? The problem we’ve all had in this industry is we have a complex technology like infrastructure. And if a customer has a problem with it we generally hire fairly talented people to pick up the phone and say, ‘Describe your problem – and I can help you troubleshoot it’. That’s a perfect place for AI to play. And with our machine learning engine, we’ve created a capability to have a fairly untrained individual talk to an application and say what might be going on here? Customer Care is a great place to apply AI and we’re in a really interesting position to do that today.”
AI and Gen AI
The ability to talk to and interrogate operational data, knowledge and systems is of course a key emerging proposition for Gen AI in network operations and customer care. Although the BT demo might be rules-based, Hamilton says Infovista has taken steps into Gen AI, albeit it is still early days.
“We’ve demonstrated and integrated it with our technology stack and now we’re starting to talk to customers to say, ‘What does this look like for you?’ Because you have all the issues with Gen AI – data access, privacy, what are you feeding it? What are you asking it? What is it giving you?”
“We don’t have it in the market yet but we’re starting to prove to ourselves that the data that gets created from things like network testing and planning – an application where you put in literally thousands of pieces of data criteria – if that data could be encapsulated and your assurance engines can sit on top of that data, in the AI world you have a beautiful thing. You have intent and you have what’s happening in real time. If you can bring those together that’s pretty interesting.
“So we’re spending a lot of time and energy exploring the possibilities that our customers can take advantage of. And, quite frankly, back to the BT example, we’re just starting to spend more time with customers to co-develop. Nobody can do this on their own. It’s experimental but we have a pretty rich set of capabilities.”
When it comes to co-development, Hamilton mentions one customer that is looking to reduce or eliminate physical drive testing, using AI to think differently about how it tests its networks.
“In benchmarking and troubleshooting I think AI in particular could help us be smarter. Can we use AI to tell us where we should be testing? In the simplest example, I get a bunch of calls from this particular part of the market – can we can we solve that problem without sending somebody out in the car to show up to do a network test”
“That’s where I again, I think we feel like we’re in a unique position because we’ve been in the assurance business for so long. We know the network testing business, and we have a really good understanding in our teams and in our technology about how these networks are designed.
“So how do we revisit the way we do work in the context of managing networks? I think that’s the most interesting thing right now because it’s the most pragmatic, moving from rules based correlation to how does your network learn, how do your operations become more efficient?”
Market moves
It says to me that the industry is starting to recognise that being able to access data in sophisticated ways is really, really important to the AI story and to the automation story in general.
Recently, in the test and assurance market, Viavi first tried to buy EXFO and then Spirent, where it was then outbid by Keysight.
Why are we seeing moves to consolidate in this market, in Hamilton’s view?
“I think what’s really happening – and this is my own opinion – is that more and more companies are starting to understand that the biggest limitations that exists on AI right now are around data. So you got to have the right data in your environment. And it’s not just AI but it’s your whole automation strategy – you can’t automate if your automation engines can’t learn, and they learn from data.
“Companies like Infovista are becoming more and more interesting in this AI play because what we do really well is collect data. We have been in the business of pulling data off networks to analyse and use in an interesting way. So when I see things happening like Spirent, that was very interesting to watch. It says to me that the industry is starting to recognise that being able to access data in sophisticated ways is really, really important to the AI story and to the automation story in general. The combination of active and passive testing and network designs network schema is critical to the advancement of AI.”
