Subex Limited ($532348)

Good morning, ladies and gentlemen. I am Madhvi, moderator for the conference call. Welcome to Subex Limited Q4 FY '26 Earnings Conference Call. [Operator Instructions] Please note that this conference is being recorded. I would now like to hand over the floor to Mr. Pratik from Investor Agency EY. Thank you, and over to you, sir.

Thank you, Madhvi. Good morning to everyone who have joined the earnings call for the quarter ended March 31, 2026. I would like to introduce the members of the management who are present for the call. Ms. Nisha Dutt, Managing Director and CEO; Mr. Sumit Kumar, CFO; Mr. Harsha Angeri Head Corporate Strategy and AI; and Mr. Ramu Akkili, Company Secretary and Compliance Officer. I would like to start the conference call by going through the safe harbor clause. Such statements in this presentation concerning our future growth prospects are forward-looking statements, which involve several risks and uncertainties that could cause actual results to differ materially from those in such forward-looking statements. These risks and uncertainties relating to these statements include, but are not limited to, fluctuations in earnings, our ability to successfully integrate acquisition, competition in our areas of business client concentration, liability for damages in our contracts, withdrawal of tax incentives, political instability, unauthorized use of intellectual property and general economic conditions affecting our industry. With this, I now hand over the call to Nisha D to take it forward. Over to you, Nisha.

Thank you, Pratik. Good morning, everyone, and thank you for joining us today. I'd like to begin with a brief reflection on my tenure here, particularly as this moment marks not just the close of the quarter, but also end of the financial year. I've been here for 3 years now. And when I came in, I came in with a clear mandate. It was to transform Subex and return it to a growth path. In this time, we have strengthened the fundamentals. We have exited noncore businesses. We have freed up capital. We have rebuilt the foundation that this company needed. Many of you have tracked this journey closely, and I believe the first phase of that transformation has been is not just a claim. I think it's a fact that's visible in the numbers and in the business that we are running today. If I were to look at this year, I would think that this was a year of transition for us. We had leadership changes, Board evolution and a very challenging macroeconomic environment. But these are not small changes. So I won't minimize any of that. But I also don't want to let it define the narrative because the narrative of FY '25 is something very different. We didn't just hold steady through the change. We actually move forward to it. Our order intake grew 24%. We released new products. We went to the market more aggressively. We won more deals. We included more AI deals than any other prior year. We also took on more AI models into production this year than ever before. On the Board, I'm pleased to share that we have also added another independent director, further deepening the experience and perspective around the -- so this is not the signs of the distracted business. This is a business that reset and kept building. And by that measure, this is one of the strongest years in the last three. And that was the original promise, and I think we have delivered it. I'm aware that we could have moved faster on the top line, and I own that. But I also want to be very unambiguous about the growth. I think that is going to be my particular focus in FY '27. On Q4, we closed the year with a 3% sequential growth. Full year FY '27 total income came in at 6% Y-o-Y. It was broadly supported by interest income. We are growing, but we are not at the rate that I would have liked us to sell. So I'll be direct any time that you ask that question of me. But one important context is important -- it's important to remember before I go deeper into the numbers. FY '26 and FY '25 are not clean like-for-like. If you remember, we exited noncore businesses last year, which means that we are comparing against a different -- so if you strip that out, the underlying picture is cleaner than the headline suggests. On profitability, our EBITDA is at 14.5% and PAT at 13.6% for the quarter, positive EBITDA in nine out of the last 10 quarters, three successive quarters of positive PAT. That consistency does not happen by accident. It reflects 3 years of genuine operational discipline, and it gives us a platform to accelerate strong, not just a baseline to defend. Our liquidity also strengthened by INR 70 crores over the year. And I want us to sit on that for a moment because this isn't a balance sheet footnote. It changes how we operate. We can now invest with conviction, move on opportunities from a position of strength and stop managing defensively. That is a materially different posture, and we intend to use it fully. And when I look at all of this together, the profitability consistency, the liquidity strength, the 24% order intake growth, the AI traction, the quality of rent and then look at where our share price sits today, it's very disappointing for me, and I'm sure it is for a lot of you that the market has not really caught up with the business. So I truly believe that the stock is materially undervalued relative to our fundamentals and our trajectory. And I think that gap will close, and we intend to close it by executing. The fundamentals are in place. The pipeline is converting. So let me walk you through what that looks like on the ground. New logo in North Africa for our enterprise asset management solution. Our competitor accounts we won back for business assurance in Middle East. In my view, one of the clearest signals where our product credibility stands today. Customers don't come back unless the product earned it. A new deal in North America for AI handset fraud solution on FraudZap and a Tier 1 operator in Africa who didn't just renew their managed services engagement, they expanded it into AI use cases. That large fund matters because it shows that the existing customers are deepening their relationship with us. And that's the leading indicator of where this business is headed. Key geographies, multiple solutions and increasingly AI at the core of it. We are also making sustained investments in Gen AI-based software development. We are embedding Gen-AI across our entire SDLC. It's to lift our productivity, accelerate releases and raise the quality bar on what we ship. So this is not just a one-off initiative. It's a structural investment in how we build and how fast we -- the compounding effect of it will show up in our ability to deliver to customers and to the market. And now on FY '27, we are 6 weeks in. The geopolitical environment is complex, and I won't pretend that we are immune to it, but we are watching it with clear eyes. But here is what I keep coming back to. The pipeline is real. The market conversations are strong as they have been. And for the first time in years, we have the balance sheet to fully back our ambitions. We are going into this year on the front foot, not hoping for growth, but we are organized and resourced to deliver it. So I believe that the reset is done, the foundation is built and FY '27 is where we will convert it. So next, I will quickly cover the consolidated financial results for Q4. All the numbers are in INR. The revenue for the quarter stood at INR 72.96 crores as against INR 70.79 crores for the previous quarter. EBITDA for the quarter was at INR 10.58 crores as against INR 9.1 crores for the previous quarter. Normalized PAT for the quarter is at INR 11.51 crores as against INR 7.68 crores for the previous quarter. And PAT for the quarter stood at INR 9.93 crores against INR 2.9 crores, including exceptional items in the previous quarter. And thank you for listening. I don't just wait for your questions. I welcome them. So let's get into it.

[Operator Instructions] The next question comes from Ajay Doshi, an individual investor. Sorry, there is no response. The next question comes from Sanjyot Khare, an individual investor.

First of all, congratulations on a good set of numbers and good operational results. And in last 5 years, I think this is one of the best result as we have been with the company for a long time. And just my question is about we are seeing a sequential growth of around 3% in the revenue. And in last 1 year, I think company has got orders more than INR 150 crores in whatever announcement was done by the company. So now is it really not converting into revenue or when it will get converted more into revenue? Is it because of whatever the geopolitical problems are happening and the orders are not getting delivered or closed for converting into revenue? And are we seeing that at least for Q1 whether the revenue will grow faster than earlier quarters?

I think it's right that we are entering the year with a strong backlog, first of all, and we also have a strong pipeline and the order conversion that has happened. But the way our contract is, first of all, most of the wins that we have had last year, they are as much for this year, but they are also for the future years, right? So we do have multiyear contracts. So you have to see it in that perspective. There are two ways to look at it. It also gives us annuity revenue as we go into the future years. So one is we are entering the year with strong backlog. The other thing is that for all the contracts that we won last year, we have started executing those. So once we complete the implementation phase of our product, our subscription revenues will kick in, and that's when you will actually see a lot of uptick. But I do expect in the future quarters as we go through this year, you will see that as we go from quarter-to-quarter, our conversion will look better and better because that's how it works because as you get to the end of the implementation cycle, your revenue starts really kicking in. So we are expecting to see an uptick from -- on the back of what we won last year.

Sure. And the second question about how this market situation now because of whatever the geopolitical things are happening and a lot of work Subex is doing in Middle East as well as the countries are impact in geopolitical country. So that is one impact and second again, the AI impact which is coming up. And a lot of these companies, the Anthropic or OpenAI are coming up with a lot of offerings which are complementing offerings by some companies. I don't know how that is impacting our company, Subex. So these two scenarios, how things are getting impacted for the Subex?

So on Middle East, the risk is actually so obviously, first of all, it's very unpredictable as all of us know that situation is changing on day-to-day basis. So it's become a little unpredictable for us and we do have quite a bit of exposure in Middle East. But that said, what we have done is that we do have a Middle East mitigation plan, right? So a lot of our delivery, for instance, have moved offshore. We have offshored our delivery already. So last time when actually the war broke out, we moved a lot of resources offshore. And we are continuing to work with our partners. They are also very supportive in Middle East for offshore delivery model. So to the extent possible, we have avoided the exposure and the work continues. We are doing it offshore instead of onshore right now. So we are also going back to a [ cop ] with protocol. So we are adopting some of actually and moving it. So in some sense, we are able to mitigate it. I think there are two parts to the risk, right? One is delivering on the projects that we already have won. The second part is there we can control it by offshoring some of it. The second part is, will you get the renewals that you were expecting? And will you still win a lot more in that market. So in that renewal, I think we seem to have a good control on the renewal part of the story because customers are still working with us. So we are very confident that we will get the renewals. On winning the new businesses there we might struggle a little bit because it depends on how fast the customers will move with allocating money to newer projects. So that's where a little bit of uncertainty comes for us as I think everyone else. So if I were to think about it, I would say that we have a sound mitigation plan. Two of the three buckets are kind of, I would say, a lot more under our control or we are able to exercise control unless it breaks out into a much bigger thing than it is right now. But currently as we are able to control that quite a bit on the Middle East exposure. On AI, I actually think of OpenAI as a tailwind actually for our business. We are what we call -- and for those of you who have been following, there is something called system of record, SOR and then there is system of intelligence, right, SOI. So what we think of our systems are -- we think of ourselves as SOR actually where we are embedded in telcos. So we are the system of record and we are strengthening that. And we deal a lot more with qualitative data, right? So what we are trying to do is trying to build an SOI layer on our system of record and trying to control that whole stack. So for us, it's been a tailwind actually. I don't see -- think of it as disruption. If anything, I think it's giving us newer tools to actually go out in the market and build the intelligence, which we were already sort of sitting on. So I don't see it as a threat right now, honest. anything we are very excited about what's coming because it's giving us productivity internally, it's helping us move faster. It's helping us build a lot more faster. So I truly think of it as a...

That's great. And my last question is about as Q2 and Q3 of last financial year, we are seeing a lot of orders value of the orders were also high. In Q4, it has gone down a little bit the orders are reduced and it's obviously some delay in decision-making or maybe what the things are happening. But how are you seeing like the demand for the services offered by Subex in other markets which are not directly impacted by the geopolitical. So are we seeing that the traction is better and the things are moving forward and things get deals closed in Q1 and Q2, more deals will get closed in Q1 and Q2?

Yes. Of course. So the deals that you saw in Q3, so what happens is that there are three types of businesses that we do, right, or the deals that we win. Some are we call them as the blockbuster deal. So some of the deals that we saw last year, for instance, the over $6 million or so, those are blockbuster deals for us. They don't happen that they don't happen every quarter. But when they happen, they come with a big number, right? On a quarterly basis and on a business as usual basis, our deal sizes are in the range of $1 million, $1.5 million. That's what you see typically. So that's the size. So anything that goes over $2 million, $3 million, it's a blockbuster deal for us. So you have to see it with that -- a little bit of that perspective. We continue to build our backlog, like I said, right, our backlog is very strong and our OI when I say OI, that's our order intake. Our order intake last year was 35% higher actually than any other year. So we have done some amazing work last year. So we do -- that gives us a strength of backlog as we enter the year. And we continue to see that momentum actually even from Middle East, I would say, not just from other geographies, even from Middle East. It's just that in Middle East, like you said, the deals are won, but sometimes what happens is the I is not dotted and the Ts are not crossed. So at the end moment, there are some contracting delays that are happening. But otherwise, the deal win is there actually. So I still think that we are -- we have a lot of momentum in the market right now.

[Operator Instructions] And the next question comes from an individual investor.

What are our margins for the three services verticals?

You mean the [ rafm ] or

We have categorized our revenue by three segments, right?

Yes, license [Technical Difficulty].

What are margins spread?

So typically license is high as you can imagine, that's typically like 100% margin, right, on license, 90%, 95%, very high margin for license. Implement margins vary from project to project depending on what is the complexity of the project. But typically, I would say implementation margins are also in excess of -- they can be in 40%, 45% range implement margins. They can even go up to 60% depending on the complexity of the project. So there is great variability in that license -- sorry, in implement component. Then MS is obviously, I would say this is high-quality revenue typically for us. So that's 5% plus on MS and AMC is 70% plus on the margin. So that's typically the margins.

Right. But then Nisha, if I were to compare apples from what you call tools as per margins, right, they are not translating into EBITDA if I take the weighted average of what categorized our revenues, right? So

... that gross margin, what you would classify as COGS, goods sold or if you were to look at what I'm talking about, these margins are typically gross margin, right? EBITDA is your off margin. So it doesn't flow directly to EBITDA. There are again expense line items below your COGS line item, which is your gross margin. And then you have some expense line item that then comes your EBITDA. So there is cost sitting below that. And that cost is typically the cost of -- just to explain it for everyone's benefit, that cost is mostly like R&D cost, which is our engineering and our GTM cost, sales, marketing cost will be our G&A, general and admin, lease expenses, all that and after that EBITDA comes. So that's why you see that our gross margins are pretty good actually. By the time it goes down to EBITDA, we lose through the line items that are kind of where you build the market and take it to market. So those line items are sitting below this gross margin. So that's the cost structure.

Right. So can I say these are the operational items? And is there a way we can further take steps to optimize those costs? Is there an opportunity there?

Absolutely --that is where our --actually a lot of our operating margin EBITDA has come from optimizing those costs because that's the only leverage I have in the P&L. So we are looking at cost, right, cost efficiency. When I say AI SDLC, I will use AI to develop my products and accelerate that actually reduces my engineering cost, which is R&D line item that reduced effectively. So that flows down to my EBITDA. So all the EBITDA improvement has happened because of this controlling the cost. So it's a direct thing. So there is a framework for turnaround that we were following. And our framework for turnaround actually has this very strong focus on what our R&D. So for instance, should R&D be benchmarked at 12% or 15% of my overall cost? What can I benchmark it. So we have created our benchmarks and we strongly operative.

So one more question regarding the margins. Is there a classification that you could give me for the margins for our old and new products? For example...

For example, -- you mean the new products like FraudZap and all that...

Yes, exactly. And our legacy products that we have been working with the clients. [indiscernible]

Products, you can imagine all the legacy products, the ones that you see in MS and AMC are all legacy products actually. So those margins have discussed. In new margin, I'm a little reluctant to tell you that because it's only -- it's competition sensitive on this call, so I...

Okay, point taken. Ballpark number taken...

Obviously much better, Abhishek, because we have built it with a lot lower cost, right?

Perfect and then that makes me happy. Another question, if I may continue. conservative and stretch targets, the question that you don't like to answer most targets for '26, '27 forward-looking statement.

Ramu is staring at me. Ramu, the company secretary is staring at me. Curiously shaking his head that no forward-look guidance.

[Technical Difficulty] you phrase it in whichever way it...

We are going -- we are entering the year with very strong backlog. We are entering the year on the front on the back of a lot of wins last year. So I am -- if I were to give any -- without taking the numbers, if I were to give any guidance, I would say that as clean as everyone, we are actually very bullish right now. So we are very bullish on this year. And I feel that by all benchmarks, we should beat everything that we done. So, we are going [Technical Difficulty].

Okay. I have three more questions, if I may - I am sorry.

The only one thing is trust the process. I think it has given us a lot of results until now. So please trust the process. It's okay.

Okay. I have three more questions. May I squeeze in or do I have to come back in the queue?

Maybe you can. I think your turn will come pretty quick. So you can come in the queue.

The next question comes from Ajay Doshi, an individual investor.

I think he has a problem with his line.

Yes, ma'am, there is no response. [Operator Instructions] The next question comes from Mehul from 40 Cents.

I am new tracking this company. I have two questions. We are into the software development.

Mehul, we are not - audio clearly actually. Your voice is breaking. I am not able to hear you what you are saying.

Is it better now?

Okay, let's see if we are able to hear you.

Is it better?

I think so.

Hello, is it better now?

I think yes.

Hello?

Yes, hello, I think.

Yes. Am I audible now?

Yes, you are audible.

Okay. Ma'am, I just want to understand what are the products which we are operating -- I mean what are the products which we are serving our customers? I understand that we are mainly in the telecom sector. But what are the products and what kind of services? I mean and also what is our split between product and service revenue? That is my first question. I have another question. Go ahead, please.

So products we mainly have three or four products. So, one is revenue assurance. Then we do fraud management solutions for telcos. So, you are right, we typically [ lead ] those telcos. We do -- revenue assurance is basically protecting revenues for telcos. Fraud management is helping them detect and fight fraud. That is where we also use a lot of AI actually. The third one is partner ecosystem and partner settlement system, which is really wholesale billing. So we do wholesale billing settlement for telcos. And the final one is enterprise asset management, which is essentially you discover assets on a telco network and you kind of reconcile that with their fixed asset registry. So that primarily, the four products. Obviously, we have embedded, AI and injected in our product portfolio. But where we see the strongest need for it is actually in fraud management with medical assurance. So fraud management also because you are seeing the fraud on your phones today, right, the WhatsApp fraud to call fraud, to link, click on the link, phishing, smishing. So a lot of fraud types are now digitally evolving because the attack vectors have increased with the number of devices that we use today. So that is primarily our portfolio. And we are a product company, we build products and we sell products and our service revenue is around the products that -- so we service only our product. We are not a general SI system integrator that goes and services any product. We do services only around our product, I hope that answers your question.

That answers -- my follow-up question is that you mentioned that mainly you are providing fraud detection solutions. So how does it work? Because let's say, I am a telecom customer with Jio, for example. And so where will your product sit? I mean, I just want to understand because I will be using Jio, but somebody calls me and then there's no control who is the guy who's calling me and what kind of fraud he can do. I mean, he can ask me for a pin. So I'm just trying to understand how does it work? I mean maybe I've given the wrong scenario, but there may be some scenarios where your solutions work. So I just want to understand that.

So fraud is an ever-evolving landscape, right? I mean one is kind of answered a little bit, but I think if you are interested in really understanding the product portfolio and what we do, then I would also say that maybe you should kind of look through some of our collateral. But in terms of very quickly, if I were to answer, I would say that our system sits deep in telco, right? And what they do is they look like -- so we can solve things like SIM-swap, Wangiri fraud. Now the fraud that you are talking about, for instance, for WhatsApp fraud, these are digital fraud that have evolved quite recently actually. This is the legacy and this is not very high. And there are solutions that are targeting solving some of these things. So we build solutions to see if they can help telcos fight some of these frauds. Now will all telcos use the same solution? No, because depending on what the telco infrastructure is, the solutions also vary. But our point is that we are helping them fight quite a bit of fraud actually in the system. So an occasional call that comes to you or a WhatsApp call, it's not to say that because somebody buys our product, there will be no fraud in that network. Fraud is evolving actually very, very fast. Fraud is a very fast evolving right from deep to account takeover. It's actually moving at a very fast pace. So this is something that we have to be constant on top of.

Right. Ma'am, how much of your -- how much does your revenue come from top 10 customers?

I think about 60% comes from our, I would say, the top tier customers.

60%, is it?

Yes.

Okay. And are your -- is it possible to know your key clients, your key customers?

I think we already disclosed it in our investor presentation actually. So our investor presentation has the logos that we are working.

Okay. I'll look into it. And last question, ma'am, because mainly whatever little I did check, I just checked Subex's numbers today itself. So I don't know much. But what I understand is that there is no promoter holding. So who is the largest shareholder in this company who are the key shareholders, I mean trying to understand.

No, that's true. We don't have promoters. We have a very diverse shareholding. And bulk of the shareholding is retail investors. So we don't have I mean our [capital ] is pretty diverse from that perspective. So it's primarily retail shareholders.

And what kind of shareholding does the management have I mean, because that is -- you guys should have skin in the game, right? Otherwise, there is kind of --

So the Board has taken cognizance of this ask that has come up repeatedly actually. So management does have ESOP. I can -- and we are trying to -- every time that the ESOP pool becomes available, we are trying to kind of increase the shareholding of management consistently. But the way management holds today is ESOP... ESOP.

The next question comes from Abhishek Kale, an individual investor.

What's the progress in onboarding our HNI investors and MFs? [Technical Difficulty]

Your voice is breaking a little bit.

Okay. Is it better now?

Yes.

Okay. So what's our progress in onboarding HNI investors? We did want ordinary capital presentation, but after that, have we received any kind of response from the HNIs or...

No, not yet. We have had some conversations, but not yet actually. I mean, I don't have them on the cap table. But what we are going to do is that we do have a plan to actually hit the road and that is this quarter. So we are going to hit the road and we are going to do start presenting because I think now we have the results with us. So we are going to hit -- so this quarter or this year, a lot of my focus is going to be to see if we can improve the cap table right now. But this quarter onwards, we are hitting the road actually -- so we are going to meet a lot of people.

Okay. So one -- I think this is more of a feedback. I think when we sign our deals, right, in the exchange announcements, we disclose what is the likely order size and the tenor for the order, right? But if we can include milestone payments, right? So this is when we see that these milestones would be met and this is the likely percentage payment that we are likely to receive that -- I mean, for us, the $6 million deal, let's take that. It was -- it's a 5-year deal for us, right? But I don't know how and when am I going to get the revenues, right? So it would make a lot of sense for investors to understand milestone-based payments as to when we receive how much. I mean ballpark is fine, but I'm sorry.

So I think that might be challenging only because contractually, we are bound by some confidentiality issues with the customers -- disclosure, we have to make sure that the customer is okay with it. And so I think contractually, I will have to check. I'm not sure if we can do that actually. Contractually, they will not agree with it. And also it's somewhere a little bit competitive information, right? How we structure our contract, how much of it is front loaded versus backloaded. So some of this falls into that sphere.

I mean without breaching the contractual obligations and without having to disclose to our competitors what we are. I mean if there is any rough idea that we can get if possible, it would be great help. And another question is on the goodwill write-off. Is that past history done? What have our auditors said on that? Because I think two quarters back you had hinted that we may have to look at the goodwill on our books, about INR 190 crores plus month. So what's happening?

So basically, it's an annual exercise. impairment exercise is an annual exercise that we do. And Q4 being the balance sheet quarter it has to happen. So that is why I had indicated earlier that it might -- we will go through that balance sheet exercise. So we have completed that and our auditors have concluded that the carrying value is fine right now and there was no need to impair any of them. So that has also given us a lot of forward and positive momentum because the goodwill value is compared against your current performance. So we also felt that there is absolutely no need to impair this right now. So I think investment has been a good outcome for us.

One last question is about our Board. complete? And have we received any initial feedback from the new members of the Board from their vast experience in this field as to how we can better our products offerings or approach on the sales side? Any inputs of that nature?

Actually, they are very, very engaged Board. So it has been an extremely engaged board. And we have received a lot of feedback. And I think this is a very different board because the ones we came in, the first thing that they asked us for was for tech deep dive. So they said that they wanted to do a tech deep dive and they had to organize sessions just taking us through our products, through the demo. They have given us feedback not just on how we can build better, but they have also given us feedback on how do we do GTM better, how do we take our products to market better. So they understood how the AI agents work. So I think in that sense, it's a very -- I mean, if I were to use the word switched on, right, they are very switched on. And we just went through our AOP and budgeting cycle for Q4 for FY '27. And in that also, they were very pushy with us, right? So all the plans that we presented, they really pushed the management to keep asking why can't you do better than this. Shareholders are waiting for some better results, why can't you do better? So there is a lot of, I would say, good push that came from them, which genuinely made us rethink our plans and we kind of represented to them. So it's been a lot of push. It's been a lot of good feedback. And I think they're also holding us accountable. But I would say more importantly, they are very fair. And so far as my expectations or my interactions go, they have been very fair as well to all of us. It's very engaged Board.

Wonderful... Wonderful. One last question. With the recent announcements which PM made wherein he has asked us most people to try working from home, right? If a significant portion of our employee group, right, temporarily at least works from home, would that impact our operations by any means because we would have had the situation in COVID era. So would that impact as to how we function as a team?

Actually, honestly, we already are working in a hybrid mode. So Subex never went back to 5 days a week in August. So we are -- yes, yes, we don't work 5 days a week. We actually work 3 days a week and 2 days people work from home. So we are already on hybrid. And we have plenty of experience managing employees and sort of employees doing their work in some sort of a hybrid mode. So that does not bother us as much as it might bother people who call employees full time back to office. But my personal philosophy in this has always been that you need to have a high-trust environment. If we can't trust our employees to actually do right by us no matter where they are sitting, then I think we have something wrong with broken in our system. So for me, I like to work in a high-trust environment. And we will navigate as it. But honestly, I don't worry too much about it. We have a very good set working with us.

The next question comes from Mahesh Kumar, an individual investor.

This quarter, we announced deal for FraudZap. Now with this deal, how many paid customers we have for FraudZap? That is first question. And second question, around three quarters back, we have made announcement team of LLM for fraud management and fraud detection. So do you have any paid customer for this team of LLM?

Sorry, I didn't get second question, Mahesh.

See, a few quarters back, I think three or four quarters back, Subex has made announcement that they have launched a product which is made of team of LLM for fraud detection.

Are you referring to FraudZap because that's what we launched actually?

No, FraudZap is a different one. There was one announcement also regarding team of LLM?

Team of LLM. Yes, got it.

So we had the announcement and some POCs going up for that. So for that particular product or solution, do you have any paid customer now?

So on FraudZap, we kind of built the product, we did POC and we launched it and we got the first paying customer all within a year. So it's early pipeline for us. So currently, if I think about it, it's paid customer. But on the back of that customer has three telcos. So it's customer to us, but at the back -- I mean, the contract actually has telcos participating in it. So if you think about it, the product is actually embedded in three cases. So that is how the...

And last quarter also you had announcement of a paid customer for a FraudZap-- and this quarter also there is a announcement. .

Yes, that is the three customers actually because what's happening is that we are working through we are working intermediary who keep onboarding newer telcos. So we had onboarded a telco, we have, for instance, three telcos. So the telcos are onboarding on to the product. So every time a new telco onboard on the product, it becomes a new customer for us. So currently from that perspective, we have three customers.

Sorry, team of LLM?

So team of LLMs, right now, actually, we are in POCs. We have actually completed the POCs. We are in conversations with at least two or three customers to make it commercial now actually in this quarter. So we should be making it commercial. The one challenge that we are running into slightly on team of LLM is the hardware challenge because it requires some GPUs. And most of the customers, as you can imagine, are a little reluctant to procure GPUs and especially in Middle East because they don't want to make any hardware investments right now. So we are -- so the product is ready, the POCs are done. They want to sign up with us but the GPU is -- the hardware is becoming a little bit constraint right now. So we are also finding a way around it. For instance, we have figured out a way that we could probably do an agent in a box kind of a model. So we are also trying to see if we can make agents work on CPU instead of GPUs, so that's the technical challenge that we are wrestling with as we speak. So that has been slightly delayed for us from that perspective. It's not because the interest is not there. It's not because the customers are not willing to pay for it, but hardware is a little slight challenge. We are challenged on that aspect.

And what will be the impact of Iran war on our business because we are focused on Gulf region. 41% of the revenue coming from that region?

We do have quite a bit of revenue coming from that region, but APAC and Europe is also very strong for us actually. So I think APAC and Europe remains steady. Americas, we are going to invest heavily into Americas this year and see if we can bring it back up. Africa also is seeing momentum. Middle East, no doubt, does have -- will have some impact. But the way we are trying to mitigate is that all the implementation, all the projects that we have already won, I just mentioned it to Sanjyot as well, that all the projects that we have already won, we are trying to offshore those projects and see if we can deliver them from here. And customers are very supportive of that given the situation. So most of the resources, we have actually moved them out of Middle East back to India, and we are now delivering from Bangalore, most of it. In terms of contract renewals, we don't see a major risk unless this blows into a much bigger war and in terms of decisions or new purchase decisions, that's where we see a slight risk of customers delaying decisions. So that may be a risk there. So that's where we are.

The next question comes from Jitendra Bhutoria, an individual investor.

So I would like to ask you about the - you have closed your Sectrio business. Now there were certain receivables for which we filed we were trying to take some steps for the recovery of that. So what is the status of Sectrio as of now? Has it been closed? All contract over, closed? And whether the receivables.

Yes, please go ahead.

Yes, the other one which was pending. And second, what is the status of receivables from the earlier contracts which we closed and we had written off some money, good amount of money.

So on sector, so there are major contracts, one is closed. In fact, we are kind of -- it's almost contracts -- the final contracts and liabilities we are closing actually. So I would, from my perspective, consider it closed almost. The second one is there are also receivables. There we have decided to take a legal route. So that's where we will go down the legal path and we will try to -- we are actually engaged -- we have engaged a lawyer and we are going to go down the legal path and see if we can do some recoveries there or arbitrate with the end customers. So those receivables will take some time because we are going to we are going down the lawsuit path on that.

Okay. So this is the one which you are talking about which we have already written off receivables or the contract...

No, the one that we have already provided for -- so we have already provided for a there is no risk to anything. We have already provided for it in the book. In fact, if anything if we are able to settle this as we -- if we are able to recover money and I'm hoping that we recover all of it, but through arbitration whatever recovery comes, we will product offset -- our provisions against the income that we book. So that's how we will offset the provision.

And my next question is regarding the revenue stuck around 70, 72 quarters from when could we expect these revenues to grow exponentially?

I think in the back some contract that we won last year in good OI, as our implementation cycle start come finishing, I would say by middle of this year and all that, as the implementation starts coming to a close, provided there are no delays the revenue uptake should start in the back of it.

So your wish and with our target of INR 100 crores per quarter when could we expect?

Inshallah.

That is the comment where the glory of Subex will lie.

No absolutely. No that's my war cry internally as well. That is my war cry to my team that we have to get this INR 100 crores absolute. No, absolutely, I think this is year, we go for it.

Are you looking for any acquisitions -- organic or how do want to proceed with the amount of cash which we have now. At least we can take some sort of a risk for our products?

Absolutely. So we do have about -- so if I were to kind of put some cash aside for our working capital, if I give myself 6 months runway and all that, I still have about INR 30 crores sitting on the balance sheet, plus we will also generate cash through the year, right? So we are wanting to go back and look at some potential ways of seeing if we can do some tuck-ins this year. So that is a discussion that we are having with the Board, within the team. So that's definitely on the card. We'll be looking at some opportunities.

So are you doing anything with on grid investment? Planning to increase or dilute or divest, what are the valuations?

On-grid, no because actually is doing very well right now. So because every year when we complete the year, we obviously look at -- we ask them for their P&L and balance sheet. So it has been really good for them. So since they are doing so well, my investment in them is doing well. There is no good reason for me to exit right now.

No, that's absolutely correct. So the company is doing well, are you planning to add some more investments into that just holding on...

No, actually, I would rather reinvest back into our portfolio. So I feel that I would rather -- there are more opportunities for me to grow my money than to put it outside -- I would rather take a bet on my team take a bet on some other team.

So at appropriate time you will be divesting in on grid.

Yes, when we need the money, honestly, today, I don't need the money, right? So I don't necessarily make the money in that sense. I think that there's a good bet for us. And we have enough cash to reinvest in our business. So let's just go back -- we have enough bets we want to take internally. We are excited about agents. We are excited about some of the things that are coming up for us. So we want to go more bullish on those areas. So we take our own bet.

So our revenues are mostly in foreign currency. So are we seeing benefit out of this rupee depreciation because our expense is mostly India-based?

So I mean, obviously, when we measure our performance, rupee depreciation helps us only because most of our contracts are in USD, foreign currency, so rupee. But because some of the geographies where we work, their currencies are also in high fluctuation. So on every contract, when we are new contracts or when we are in negotiations with new customers, one of the points that's coming for us is that most of the customers now want to work in their currency, which we don't want because obviously, the exposure for us is high. If the other currencies also equally are fluctuating, then I cannot take that currency exposure. So the sticky point becomes often for us that do they have the USD to give us or not? Can they do dollar contract? So that is also slightly delaying some contract negotiations and all that, but we are sticking to our ground. For us, I think we -- to the extent possible, we are only doing GBP, USD and Euro contracts.

The next question comes from Ashish Koti, an individual investor.

Can you hear me? Yes, Ashish. I have a question basically with regards to our focus majorly on the fraud management in telecom with more and more consolidation happening and with Starlink and where competition would get more severe for smaller players, how exactly telecom would be enough for us?

So see, we are actually -- so obviously, telco is bread and butter -- so we are deeply embedded. We have -- I would say that we have around 125, 130 customers, active customers. We have very low churn. I have 80% of all Tier 1 customers. So there is a lot of -- so that's where we have a lot of strength. But that does not prevent us from looking outside. It does not tell us that we should not look outside. There are sectors or areas where our revenue assurance and fraud management might be horizontally scalable. And this is where we are actually looking at seeing if we can get into this area. But on your question on Starlink, even our customers are actually signing life of Starlink. So telcos are also investing aggressively and getting into hybrid space right now. So telcos are not the telcos of old school telcos any more. So they are also taking a lot of bets and getting out. So I think in that sense it gives us a horizontal expansion opportunity, but more than that we are also aggressively looking at some very hybrid areas right now which are I would say, high growth, high tariff. So, we have identified few areas that we are starting to take a bet on that. So, we are going to start experimenting with some of those areas, take our product, look for fitment. So you are right. I mean telco is not the only play that we will have. So we are also hoping to cautiously expand into something.

Would you be able to throw some light on areas which you are exploring is one. Second within data centers, our whether on electricity side, whether we can kind of port our fraud management system has any role to play and whether on electricity side, whether we can kind of port fraud management system?

So data centers are definitely a good place. That's because data centers not just need fraud management. Actually, I would think revenue assurance is a bigger play in data centers, honestly. So I think in that sense, data centers are something that we are starting to look at carefully because it's also a very high-growth industry right now, right? It's one of the highest growing. It's almost -- data center is a new infrastructure, right? -- when telcos Subex grew last time, it grew on the back of telcos expanding massively. And we now see the same movement happening around data centers. So it's building a new infrastructure altogether. And I think in that sense, it becomes a very strong consideration for us, In terms of electricity and all that, personally, we are little bit reluctant to go into utilities, pure utilities kind of space because receivables and you know that attrition cycles are very long, they are very close and receivables and other things become problematic with utilities. So those are not areas. My product can possibly work in those areas, but utilities are something that we actively stay away from water, electricity etc.

And our revenues are totally 100% on subscription model only?

No, it's not.

It's a traditional thing where it's a license, implementation and support. So it's not 100% subscription?

No. But in terms of sales, I'm not talking about operation and maintenance. I'm talking purely in terms of sales. All our revenue is in terms of subscription model or is it also still?

No, we do sell and we do the license, right? So we sell our product, we recognize some license revenue. And then we do implementation of the product, we get implementation revenue. And that's when after the implementation is over and customers achieve what we call go live, which is our product is live and working with their system, that's when it moves into what we call support contracts, either that's the subscription part of it. So it moves into support, which is -- it can be an AMC, it can be an additional managed service. So support contracts began after that. So that's the subscription piece of it. So there are three distinct pieces to how the revenue moves for customer.

Then if revenue mix can be broken down into different heads then...

It is. In investor deck that is how it's broken up and I was telling Abhishek also earlier on the call, that's exactly how it's broken up. You will find it as license implementation, you'll find MS and you'll find AMC. So MS AMC is the subscription. If you want to think of it as subscription MS AMC subscription bit. The license implementation is a product part of it. So if you - if that.

The next question comes from Mahesh Kumar, an individual investor.

Yes, Nisha, in multiple past investor calls, you have been mentioning that your first priority is to take the ship out of storm. So as on today, is the ship out of storm? That is the first question.

Yes.

And second question, if the ship is out of storm, what are the new growth engines that you are trying to put in place, which will give at least more than 10% growth? And third one, since you are focusing on data center, why the data center companies who are also software companies, also services companies will use your product rather than building that in-house? These are the questions.

Fair questions. So ship out of the storm, I think that every time we think we get out of the storm, a new one looms. But I would say that in terms of overall direction, I actually, like I said earlier when I was the initial opening remarks, I think that is out of the storm. In that sense, the first part of transformation, I think we have delivered. So now our focus has to be on growth because -- and we have a balance sheet actually. So I think in that sense, the fundamentals are back in place, right? So when fundamentals are in place, I can confidently say that we are on our way now. So that gives me confidence that we are over the work. So second part of your question is on -- I think you asked about data centers. So data centers are...

No, growth engines. What are the growth engines?

So our growth engines broadly come into, I would think of it as four categories, right? So there is a traditional telco that remains for us. Telco themselves are becoming tech. So I think like I was saying that there are horizontal expansions with JV being done by Starlink and all that. So that gives us a horizontal expansion. There is TAM there. Then there is a new thing called Agentic in-house and [ MS replacement ], right? So wherever they have MS fee, all telcos have a lot of managed services fees. So wherever they have managed services, we can actually go and place Agent AI and see if we can replace or reduce the managed services fee. That is a very -- the third is the new age leakage vectors, right? So FraudZap and all that control those new age leakage like social engineering, A2P SMS fraud, account takeover. So we have built a suite of products that can now address the new leakage vectors. And the final one is telco adjacencies. That is where the mobile money fraud, MVNO, IoT assurance, telco fintech data centers, all this comes into telco adjacencies that we are strongly looking at. So if I actually look at the TAM, this is 4.3 -- so 4.3 billion of TAM that did not exist even 2 years ago. So a new TAM market has opened up completely for us. So I actually think that the play for us is quite big. And these are the exact levers that we are going to anchor our business off. And all of quite fast growing, right? So this is -- none of this is a slow growing market ever.

What about data center? Why the data center company will use your product when they are themselves having huge software teams to build it?

I think that goes without saying, right, telco, one thing is you should remember that wherever our products are telcos for scale, right? So we shouldn't forget that there is one is scale. One of the telcos that we work with, and this is not an isolated example. We process like 120 billion records a day. So software is not just a software. It has to work at a certain scale. And we have production grade software that works at a very huge volume and scale. So I think that's one good thing that we have going for us. The second is on some of the products like revenue assurance and all that, even telcos huge team. Why do they still use us? It's because there is a maker checker system, right? You cannot actually make your revenues and also check it. It's like auditing pretty much, right? We use auditors because they cannot -- I cannot report my revenues directly to you unless an auditor puts a field and says I've seen it externally and it's true. So a lot of our software is also, I would say, an audit kind of a compliance kind of a requirement, right? And the third piece is domain knowledge. That is very specific to solving fraud. The domain knowledge doesn't build overnight actually. So we have our team for 30 years, right? We have seen patterns across 30 years across telcos. No individual telcos can beat us when it comes to the I know patterns across 100-plus telcos. And anyone to come and beat that with -- they cannot. And that's exactly the reason why people continue to buy us. And I think it is these are the kind of skillset -- no, I'm coming to the data center question as well because that is the kind of skill set that has put the test of time in telcos. And we believe a lot of it is actually horizontally scalable because it's the same stuff when you actually go into data centers. They still need revenue assurance that a maker checker auditor kind of thing, right? So they will still continue to buy from outside. We have to see if we are the people who actually go and build this game. So I mean that's at least our thought process. And it's not about code writing, right? If you could just win this by having a team of developers, then I think everyone could just enforce all this. But that's not what it is. Patents evolve over time. You need someone who's working with across telcos, across data centers and can bring those patents into the products and embed it and put it in your system. So that's where we think our play will be. Like again, I said that we have not honestly gone into the market on data centers yet. This is one of the adjacencies that we are considering strongly. For instance, because we can solve fraud problems, people are starting to use it for mobile money now. Why? Because I can solve fraud actually essentially and mobile money sees a lot of fraud. So we are already seeing that our products are horizontally scalable. So that's the best actually. Data centers, again, I don't want to give anybody an impression here that we are working with data centers and all that. We are starting to consider it because that's definitely a very strong growing sector and it's building, I would say, the next layer of infrastructure in the world. So it's very natural for us also to see if there is a play for us. And that's something we are strongly considering right now.

But do you still want to go back into IoT assurance again because we have exited that business recently?

We actually never did IoT assurance. Sectrio was industrial security. It was never actually an IoT assurance business in the first place. So I think there seems to be some confusion. And IoT assurance is actually a set of controls which are RA controls, those are revenue assurance controls for existing telcos. So we already do that and we are wanting to expand that business within our existing telcos. So IoT assurance is one of the assurance businesses for our existing RA portfolio actually. It's not a different product line.

See, Subex is very proud of its domain knowledge, it is having the pattern it has learned over 30 years' time. So if that is the case, I have a fundamental question why we are not building a Mythos type of model ourselves?

That's a you have to give me - you guys have to believe in us more. I want Anthropic level salary. 100 million is what we need to build Mythos, so that's what it is.

That cost is in the U.S. India will be much less than that? Tech Mahindra MD told I will build that ChatGPT type of thing in $1 million?

See. There is a okay.

So Mahesh, [indiscernible] But what we are doing with our domain knowledge. I mean, I won't take Mythos example, but I will take the Claude, Skills and Plugins, I am sure, you are familiar with that. So, in our context, our agents are equipped with proprietary skills which is our domain knowledge. I mean, you can kind of think about it like that. So we can provide skills there was an earlier question about LLM. So LLM, the Claude or OpenAI can build, but the proprietary skill we bring. That is why our agent can investigate fraud much better than any human or any other agent because we have all these patterns which we bring the vendor in the form of a skill.

See that is what we are using AI to build services or solutions that approach. But when we say...

I mean, we think of agent as a product because it's positioned as an independent product stack for which we sell and we kind of get again license revenues like was already discussed.

Because what is happening, nobody think Subex is an AI-first company. This is the impression in the market today. See, there is gung-ho about AI-AI across the world, but nobody is considering Subex as an AI-first company. So I want to bring that impression by such development that it is AI company.

It's a fair point. It's a fair point. And I think I will take this as a feedback. I think we have a lot more work to do on messaging them, and we will make sure that we do a lot more work on it. So it's a fair point. I kind of take your feedback on that -- but that's not to say we are not doing the work, Mahesh. Because sometimes...

See basic technology development. Now if you see this Anthropic and all those, they are building the technology, building the model is the technology. the Indian companies are doing using that model, we are building the services and solution. So if Ruchir Sharma, who is an investor, he told India has lost AI-first. If you see his interview, every year, he gives what are the trends we are likely to see in 2026. He told AI bus India has missed because Indians are not innovators. They are good followers.

But I think I kind of have a slight disagreement with that position because it's like saying that deep research is the only research worth doing. A lot of India's growth has also happened on the back of applied research, right? You don't need to be a deep researcher. The deep research requires deep pockets and deep money, which I don't think Indian markets have today, maybe that kind of muscle. The muscle is still sitting in Silicon Valley. So they are doing deep research. They are coming up with fundamental models. But that does not stop us from actually taking what they have built and applying it to our domain. So I would say that we are the applied research companies, right? And there is nothing wrong with it. There is actually -- it's hard work. It's difficult work to do even applied research. And I think that if India and companies like Subex can actually do that well, then we would have also won a major project actually. So I almost think of it like that, that our AI can work at magic skills. So why should we kind of think that we have lost the race I actually think the race is just beginning now.

You know what happens.

The forefront of it will get the win. So I have a slightly different perspective from Ruchir on this.

No, what happened, you were saying the hardware GPU is the issue for team of LLM. Okay. So the out-of-box thinking would have been why can't we use Neuromorphic to implement all this team of LLM. Your cost will come down drastically and you will have a patented solution.

Why can't we use what, sorry?

Neuromorphic processor. See, we are actually looking at, in fact, boxing it into very much a laptop right now, so that we can just take our laptop, walk into a customer site, plug in and start playing from there. So we are also -- so it's not that we are not thinking through that process. When we get that roadblock of GPUs, we are not going to not push our agents in the market because some people don't have the wherewithal to buy GPUs, right? So I'm going to...

Will be challenged by the Neuromorphic because Neuromorphic is basically AI at age. See, it is a.

I will look into what you're saying definitely. But my point is that it's a - yes, it's a neuro structure, right? So my side is that we are not sitting with the challenges. We are actually vetting through it and seeing that can we make our agents to work on CPUs.

AI agent with laptop is a traditional method of solving the problem. If you want a innovative solution, you should look for new competing architecture like Neuromorphic.

So Mahesh, [indiscernible] taking into market today, the infrastructure that is available with telcos, obviously, they do not have these kind of computing. They are aware.

But you are trying to put in your own box. You are trying to put in your own box.

No, the reality is that our agent requires a certain IQ level and that IQ level, no agent works on CPU, so the fact that we are cracking it to make it work on a CPU is the kind of to unblock this GPU hardware challenge itself I think will be a big win for us, especially if we want to realize revenue in the near term. Because CPU is something our customers have access, have subsequent processes and so we can start actually rolling out a lot more agent. So, while we look for different architecture which we do but I think the reason why we are going on CPU is primarily because it is something available and can be commercial.

Neuromorphic is also available. There are companies who provide this hardware.

Okay, I will actually ask my team to kind of look at it.

That would be the last question for the day. Now I hand over the floor to Ms. Nisha Dutt for closing comments.

Thank you for all the questions. I mean again, I think we have had a good year. Like I said earlier, we are looking to do better this year. We are -- at least the team is very bullish on the year that's coming ahead. So thank you for staying with us. I'm hoping that you stick for the journey. I would like -- I always think that the best days are still ahead of us. So thank you for your patience and your confidence, and I look forward to speaking to you again during AGM and then after that in the investor call. So thank you, and we'll be in touch. Thank you.

Thank you, ma'am. Ladies and gentlemen, this concludes your conference for today. Thank you for your participation and for using [ Do Saba's ] conference call service. You may disconnect your lines now. Thank you, and have a pleasant day.