SuperOps is the all-in-one IT management platform powered by agentic AI, designed to help MSPs and IT teams work smarter, resolve issues faster, and scale seamlessly without complexity. With features like AI-driven automation, proactive monitoring, and a sleek mobile app, SuperOps boosts profitability, operational efficiency, and customer satisfaction—all from one powerful platform.
In this episode, we speak with Arvind Parthiban the innovative ways AI is transforming managed services. SuperOps has launched a unique “Agentic AI” campaign, featuring billboards in Times Square and a drone show at DEF CON, to promote the development of autonomous AI agents. These agents are designed to perform tasks independently, enhancing efficiency for managed service providers (MSPs). The conversation highlights the distinction between true autonomous AI and mere marketing buzz, emphasizing the importance of community-driven development and transparency in AI adoption. Arvind shares insights on the potential of AI to improve business efficiency and the importance of educating users about AI’s capabilities and safety.
Watch the podcast here:
Listen to audio only here:
Learn more about SuperOps.
Interested in appearing on the SourceForge Podcast? Contact us here.
Show Notes
Takeaways
- SuperOps is launching a campaign to raise awareness about Agentic AI.
- Agentic AI is about creating autonomous systems that can perform tasks without human intervention.
- There are two types of Agentic AI: human-assisted and fully autonomous.
- AI can improve efficiency by automating repetitive tasks like ticket dispatching.
- Community engagement is crucial for educating users about AI.
- Transparency and security are essential for building trust in AI systems.
- SuperOps is focused on integrating AI with existing tools for better efficiency.
- The future of AI marketplaces is promising, with potential for various AI agents.
- Education on AI use cases is necessary for user adoption.
- Cautious rollout of AI technologies is important for user confidence.
Chapters
00:00 – Introduction to Agentic AI and SuperOps
02:43 – Understanding Agentic AI: Definitions and Distinctions
06:25 – The Role of Human Oversight in AI
07:50 – Building Trust in AI: Education and Community Engagement
08:22 – Security and Safeguards in AI Systems
09:01 – The Future of AI Agent Marketplaces
11:35 – Long-term Success Factors for AI Agents
Transcript
Beau Hamilton (00:01.102)
Hello everyone. And welcome to the SourceForge Podcast. Thank you for joining us today. I’m your host, Beau Hamilton, Senior Editor and Multimedia Producer here at source forage, the world’s most visited software comparison site where B2B software buyers compare and find business software solutions. have a very exciting episode today with Arvind Parthiban CEO of SuperOps who joins us to unpack out AI is reshaping the world of managed services. So SuperOps just kicked off what they’re calling a first of its kind, Agentic AI campaign.
And I think it’s pretty wild. They’ve got billboards in Times Square. They have a drone show at DEF CON and a 100,000 hackathon inviting managed service providers and developers to build real world AI agents, which if you’re unfamiliar, they’re basically systems that exhibit autonomous behavior that can take on tasks themselves without waiting for the user. And they’re saying it’s not just another tech launch. They’re saying it’s a movement. It’s one that’s crowdsourced. It’s community led.
And it’s laser focused on giving MSPs the tools to co-create the next generation of autonomous agents. So I’m really excited to see what’s coming down the pipeline. And I’m excited to talk about this initiative. So let’s not waste any time. Let’s get right into it. Let me introduce Arvind. Arvind, welcome to the podcast. Thank you for joining us.
Arvind Parthaban (01:13.168)
Thank you so much, Will. I’m really excited and it’s an honor.
Beau Hamilton (01:17.142)
Absolutely. First of all, just want to say kudos to you and your team for this insane marketing push. I’ve been following you over on LinkedIn, all your updates and seeing the agent missing billboards and the trucks all around the country. It’s pretty neat. And it’s just, you know, it’s exciting to see kind of the next frontier, right, of AI and how it’s evolving right before our very eyes.
Arvind Parthaban (01:40.028)
Yeah, thank you so much. But let me just take a moment to explain what we are trying to do here, right? So there’s been a lot of noise about agent-dk-i. Everybody, every industry, every SaaS founder, every industry is talking about agent-dk-i. And it’s been a while. And what I noticed is a lot of them are confusing what agent-dk-i is. Old workflows are being renamed to agent-dk-i. It’s just a simple thing of doing manual work, grant work autonomous. It’s not replacing or doing something really big, meaningful. Everybody wants to wave, but we are not doing it right. So that’s why the campaign is about, agent is missing, let the real agent please stand up. That’s the idea behind it. We want to bring some awareness. As a community, we need to learn what to build, what is agent-y AI, how it’s going to change our life. So that’s the whole campaign is about and a little bit community movement with awareness is what we are trying to do.
Beau Hamilton (02:44.78)
Right, that’s a good distinction there, because I’ve seen a lot of, and I’ve heard from a lot of companies talk about just the marketing side of things. They’re capitalizing on the buzzword of Agentic AI, but it’s really not real Agentic AI. It’s not real autonomous behavior. So I think it’s neat to see that you’re actually, this is real autonomous services being integrated into your systems, right?
Arvind Parthaban (03:07.9)
Exactly, and we trying to commoditize it. So that’s the beauty of it. It’s not just we as a vendor, we are trying to say this is what Agentic AI means and thrust it upon the users. We are welcoming users to help us build and collectively figure out this.
Beau Hamilton (03:13.198)
Hmm.
Beau Hamilton (03:26.638)
Now for you guys over at SuperOps, essentially the agents are installed on your clients’ devices and they work to establish sort of a deeper connection between the admin and the endpoints in your clients’ network, right? So it’s essentially designed to take like a proactive, autonomous approach or action rather than just like passively respond to commands from the engineer or whoever it may be, right?
Arvind Parthaban (03:52.636)
Correct. So there are two kinds of agent AI we are trying to coin. Everything is new. There are human assisted agent AI. There are independent autonomous agent AI. There can be small, I’ll give an example, like a ticket dispatcher. So in the past, we used to have one of the technicians read through all the email tickets and then assign it to the right technician based on their expertise, their location, and things like that.
Now AI can do that. We have a feature which can automatically assign the ticket based on the expertise of the tech, pass tickets, and location, and all different criteria. But somewhere we need to also have human to make sure that it is doing well. And it’s not going to completely take it away from human. is human assisted. But the real agent of AI is where there is no human involved, where it can automatically solve certain tickets, L1, deflect it, and take independent decisions. So that is the real Agentic AI. So there are two folds, and we have started building these first Agentic AI where it is human assisted. It does 90% of the job. But somewhere we need to have a check before we completely give that autonomy.
Beau Hamilton (05:03.532)
Yeah, can you can you talk about some more? what are some of the specific examples where Agentic AI has improved the things like the user experience, but also the business efficiency of your clients?
Arvind Parthaban (05:14.798)
Yeah, so let’s take Knowledge Base, for example, right? So it requires a lot of tickets and properly documented solutions. So once you have all this in the Knowledge Base, if someone is asking for a question, if it is a repeated ticket, like a ticket has, this incident has happened again and again, the agent can actually scroll through all your tickets, your Knowledge Base, and tell.
These are the things you need to try, or it can automatically go and try it if it’s integrated, and then solve L1 decades. I’m saying this is something which human is not involved. It can be a simple thing like, hey, how can I update my virus scanner? And this is how it should be done. And you can figure it with a script, and you can do it by insert. This is what is happening now. But to take smart decisions, it requires more and more data. We need to train it. So that’s the phase we are working on.
But right now, there is some adoption. People are using this, but it is not replacing humans yet.
Beau Hamilton (06:16.706)
Okay, it’s not replacing humans yet, but it’s, know, that has the potential not, and I wouldn’t even say replacing humans, you’re just essentially freeing up, you know, roles to do other things, right? Now, just like with any new technology, it can take some time to get adopted for various reasons. There’s lot of factors involved, but you need to educate people on kind of the use cases, the upsides of the technology, but also especially with this AI, Agentic AI technology, you need to reassure them that it’s safe, right? Especially when you’re dealing with automated systems that take action on the user’s behalf. How do you educate customers on trusting AI agents to take meaningful action autonomously?
Arvind Parthaban (07:01.98)
This is a very interesting transformation, if you ask me about. So what we need to do is it’s a collective thing. So community is the only answer. We are discussing this. We are meeting up with lot of MSPs. We are doing a lot of educational campaigns. As we speak, we had a cab meeting yesterday in New York. We are talking to folks on how to adopt first. Collectively, it’s community is the only answer. Like, we have to do it together.
The awareness just not just comes from us as vendors, it’s thought leaders, the MSPs who are users, the challenges they are facing. So this is like we need to do this together. And there’s a lot of noise about it, but it’s about being transparent and having the clarity of what we are doing. We can get carried away assuming a lot of things, but it is very important to have a very, very close feedback loop with the customers on what we are trying to solve. End of the day, everything boils down to how can we make it faster, better, and cheaper, right? How can MSB business be efficient? So if it’s solving that, I think we have a win.
Beau Hamilton (08:04.44)
Now we hear a lot about various chat bots going kind of rogue and not necessarily following the prompt engineer’s direction. And obviously this is a different technology than some of the agentic tools and services you have. But I’m curious just what kind of safeguards you have in place to make sure these systems don’t go off the rails, so to speak.
Arvind Parthaban (08:30.46)
No, is very, very important. Security is very important. Data is very, important, especially like we are talking about working with multiple tools. And whenever it comes to Agentic AI, it’s not just on soup props. It’s going to integrate with multiple tools. It’s going to have access of a lot of tools and information, like what if somebody can actually ask for a prompt and get four passwords. These are things which is very, very important. So the way we are approaching it is we will only integrate and call the applications get the data, it is going to follow all those data center things like GDPR and data being within the US and security and governance is one of the biggest concern. And we need to take care of some kind of access we are providing the agents to actually pull, right? it should not be, data should not be stored. It only should be able to access and call and not completely take the data, right? Like that’s how we are trying to build. We are reviewing every code which is being built and then only up going like think of it like an app store, Apple App Store, we are learning a tight ship on data and privacy.
Beau Hamilton (09:34.638)
That’s good. Imagine that’s one of the biggest hesitations you hear about. Is it the security privacy aspect?
Arvind Parthaban (09:45.242)
Yes, it is not just for our industry. I think AI governance is the biggest concern for everyone here. It is something which we need to fix and learn as we grow. It is not something we know right now. And we should be very, very cautious on what kind of access we provide.
Beau Hamilton (10:03.426)
Right. Well, I want to talk about the marketplace you mentioned. So I know SuperOps has a marketplace with integrations already. You can plug into the SuperOps platform. But as far as I know, there’s no dedicated AI agent marketplace just yet. your platform, I guess let me rephrase this, if your platform could offer AI agents via a marketplace, what kind of agents do you think would your users be most interesting, are interested in and willing to try out?
Arvind Parthaban (10:34.202)
Yeah, so we do have a marketplace already, and we also integrate with Backseat, where we provide access to all the applications in the industry. I think there is no choice. Moving forward, agents are the future. And how we are integrating with lot of apps, agents are similar to that. We are going to allow agents to be used from our marketplace. It’s not just apps which is going to be built on top of SuperOps. It can be independent apps like a marketing app or how to get a Google review for your website. It can be apps which can be used on top of our third party vendors, even competitors. So we want to allow this because it has opened up a new ballgame altogether. Agents is not going to restrict you to only your platform. It’s going to help in multiples. And we are going to allow any app which is helping MSPs to make their business better.
Beau Hamilton (11:30.978)
Awesome. So OK, just to clarify, so you do have a Agentic AI marketplace already? Not yet? OK. OK.
Arvind Parthaban (11:35.62)
Not yet. yet. We have agent AI bots on top of SuperOps right now. We have not launched any market.
Beau Hamilton (11:46.222)
Because yeah, we’ve talked about AI Monica with SuperOps in a previous episode. So it’s being built out, I guess.
Arvind Parthaban (11:57.178)
No, Monica AI is like our engine. It’s like Watson IBM. It not just as an agent, it does a lot of things, Summarizing tickets. So it’s like, it’s like Jarvis, you talk to it. It helps. It’s like a co-pilot helps technicians, admins, users to do things better. You can just ask Monica something like summarize this ticket for me. And it’s going to give you information. Like you want to report about your client. It’s going to give you the information. Instead of you clicking and searching, you just talk to Monica and it’s going to have, it’s like Jarvis for MSPs.
Beau Hamilton (12:28.224)
OK, gotcha. OK, well, it’s exciting to see, to talk about this potential, this marketplace that’s coming. I know that when OpenAI announced their marketplace with ChatGPT, I was just super excited. And just seeing all the various integrations with Canva and all these different tools is just lot of fun just for the end user to play around with, if nothing else. And obviously, there’s some real use cases and efficiency gains too with the integrations. So I’m excited to see what you guys come up with in the future.
And I think one of the things you mentioned was the automated support ticketing agent. I think that would be super useful to continue to see it built out because that’s such a core component for MSPs. And again, it goes back to just having to really get that education out there about how reliable it is and how you can trust it before it totally frees up like an engineer’s tasks to do other things, right? Can you again, just talk about some of the importance of transparency, things like data control and agent accountability in this marketplace driven world you’re ushering us into.
Arvind Parthaban (13:49.02)
So think about this in a different industry. Let’s take Cursor, for example. What did Cursor do to engineers? It’s not actually replacing engineers. It’s only making an engineer’s life much better. A smarter engineer will be able to code much faster. Things which took three months to develop and launch, a smart engineer will do it in less than two days now. It’s only made a smart engineer much smarter. That’s exactly how I would put Agentic AI in the MSP world, an MSP who can actually use AI can be much more productive and efficient. You don’t need so many people to do so much. You can do it with a leaner team. And instead of hiring more people, you can deploy agents, and you can actually get things done and be much more productive and efficient and profitable.
So this is the crux of it. So the way we need to bring in AI into it is to start with, where am I spending my time? Where is my? Most of our repeated activities are where things are like very documented. It’s grant, I had to do it. Can mission do it better? Like agents do it better, right? Like these are the things which where we will start where it will have limited access. The agent will only go and do the things again and again and again like resetting passwords or things like that. But still human run MSP site.
The word I use now, think it’s catching up in a lot of our community is, AI won’t replace MSPs. MSPs with AI will replace those who don’t. This is pretty much how the industry.
Beau Hamilton (15:28.504)
Yeah. Yeah. If you’re not, if you’re not taking advantage of these tools, you’re just going to be left in the dust. Yeah. I keep telling this with all my friends who aren’t using any sort of AI right now. and whatever, you know, facet of life there or use case, I’m like, you’ve got to be using it in some way. I think you’re just, cause other people are, and, you’re just going to be kind of left behind. but…
Arvind Parthaban (15:51.536)
Yeah, that’s awareness we are trying to do. If you see, we actually posted billboards everywhere in Tampa, in Miami, in Santa Clara, in New York. We are trying to say that the real agent is missing. So just to reinforce saying that we need to be little bit more educated about it, not jump the gun and try to rush to market with half-baked agents. So we wanted to make a point, take a stance for the community with this campaign.
Beau Hamilton (16:22.444)
Yeah, I think a cautious approach is definitely wise, is the marketing push to educate users on the real meaningful use cases and then reassuring the data control aspect. I imagine you have logs of the actions that are conducted, which you can audit. And then maybe even there could be things like a suggest-only mode, where agents can only make suggestions before they’re rolled out. Things like that, I imagine, are probably pretty, pretty useful.
Now, looking at the big picture, what do you think will drive long-term success in an open AI agent marketplace? Is it monetization? Is it the community? Or is it something else?
Arvind Parthaban (17:14.618)
Initially, I think the first success would be to adopt AI. There has been a lot of window shopping. Folks are curious, but also scared. So adoption would be the first step, and awareness with community. We need to be very aware of what we are getting into, what to be built, how is it going to help us. As a community, to be honest, nobody knows the answer. Nobody knows how this is going to change the industry. Every single day, the AI evolution is happening so fast, right? We as a community should have a place where we need to discuss, decide how to transform, how to use AI. And my answer would be community-driven AI adoption would be the future.
Beau Hamilton (17:56.11)
Yeah, think just again, just kind of educating the users and the populace and being like, showing them what you can do with it, I think is just a huge step because I think a lot of times you get, it’s almost like you have these features rolled out before users are even asking for them in a sense. But once they start using them and start playing with them, they’re like, this is completely changed how I work and operate. So yeah, I think that’s very true. Well,
Arvind Parthaban (18:22.556)
And also the industry is so little fragmented. When I say fragmented, we are used to do certain things 30, 40 years. Like we are talking about an age-old industry. It’s been there for almost 40 years now. And change is not an easy thing. There is a human element involved where we getting used to the new way of doing things is something which will take time.
Beau Hamilton (18:45.006)
Well, you know, again, cautious rollout and I think, you know, I think you guys are on the right path and you have a good, you know, kind of a timeline with how you’re approaching this. And again, I’m just excited to see what’s coming and I’m even kind of more bullish on Agentic AI after this conversation. So thank you for that. Now for those, for those interested in learning more about this campaign to get, or maybe they want to get in touch with you and your team, where should they go? Where should you send them?
Arvind Parthaban (19:14.842)
No, it’s superops.com slash agent. We have all the details. For lucky few folks, we are going to take them to Vegas for the huge unveil of this campaign on August 8. So we just launched a teaser and wait up for DEF CON on August 8 and join this initiative. And you might be lucky to get to Vegas and watch this in live.
Beau Hamilton (19:38.304)
Awesome. All right, superops.com slash agent. and you know, just gonna plug your LinkedIn. It’s like, it’s a fun follow and seeing all these, all the kind of reposts of, some of the, the teaser videos of this, the agent missing, promo is, is, is always just, it’s just fun. I love it. thank you, Arvin. It’s been a pleasure. Let’s have you back one of these days too, to just get some updates on, everything we talked about.
Arvind Parthaban (19:56.176)
Thank you. Absolutely, anytime name it I’ll be there.
Beau Hamilton (20:06.158)
Sounds good. All right, that’s Arvind Parthaban, CEO of SuperOps. Thank you all for listening to the SourceForge Podcast. I’m your host, Beau Hamilton. Make sure to subscribe to stay up to date with all of our upcoming B2B software related podcasts. I will talk to you in the next one.