Cofounder, Fractal
A Strong Q4 For Fractal Analytics | FY27 Revenue Will Be Better Than FY26, Says Company | CNBC TV18
#4QWithCNBCTV18 | Fractal Analytics Reports A Strong Q4FY26
Executive VC, Srikanth Velamakanni to CNBC-TV18
Expect R&D to go up to 10% of revenue while improving the gross margin
OpenAI’s recent move an enormous validation for our products
#fractalanalytics #2026results #2026earnings #earningsgrowth #revenuegrowth #margingrowth #profitgrowth #ebitdamargingrowth #2026growth #cnbctv18 #cnbctv18market #businessnews #businessnewstoday #businessnewsinenglish #sharemarkettoday
🔴CNBC TV18 LIVE TV: https://youtube.com/live/P857H4ej-MQ
SUBSCRIBE to our Channel: https://bit.ly/3nvEcxf
--------------...
About Srikanth Velamakanni
Srikanth Velamakanni, co-founder and group CEO of Fractal Analytics, discussed the company's Q1 FY27 results in July 2026, reporting a 92% year-on-year increase in net profit and 20% revenue growth. He attributed the performance to strong demand for AI services, particularly in the BFSI and healthcare sectors, and said the company's margin expansion was driven by operating leverage. Velamakanni described the quarter-on-quarter profit decline as a seasonal blip due to salary increases and campus hiring, and stated that the TMT vertical's 22% year-on-year decline was linked to clients shifting from AI operational expenditure to capital expenditure. He expressed confidence that TMT would return to sequential growth in the following quarter. Velamakanni has stated that Fractal's 20% growth rate is below the company's ambition, and that he would be satisfied with growth exceeding 30% year-over-year. He described Fractal as an "AI first company" that uses classical AI, generative AI, and agentic AI to solve business problems. He has called on Indian IT services companies to increase their R&D spending from roughly 1% of revenue to at least 3%, noting that Fractal spends about 7% of revenue on R&D. In a May 2026 address at the NASSCOM GCC Summit, Velamakanni argued that the AI industry is not in a bubble, citing real progress on benchmarks and a major platform shift, and said the "best days of tech" are ahead. He also predicted that global tech spending growth would remain low single-digit for the current year but would begin to expand within 12 to 18 months.
Transcript (27 segments)
R
Reema0:10
Welcome back. Let's talk about Fractal Analytics. Reported a good set of Q4 numbers. The profits have gone up 14% quarter-on-quarter. They've doubled on a year-on-year basis and margins have expanded for the company. Srikanth Velamakanni, co-founder, executive vice chairman of Fractal Analytics is now joining in on the show. Srikanth, morning, Reema here. So it's been a good quarter. It's been a good year. For the full year, you've ended with 19% topline growth, slightly lower than the 20% plus run rate that you were enjoying. Now you flagged off that there was pressure in the TMT clients, and that led to a 19% drop in revenues for that particular vertical. Is that the only reason, had growth been higher, and what can you guide on that particular TMT client or clients?
S
Srikanth Velamakanni0:53
Thank you, Reema. We reported pretty good numbers overall, 19% growth for the year and significant expansion in gross margins and overall profitability as well, which is one of the things that I think you asked much of times before we went public or during the time we were going public as to how in a stable way can we make money all the time. And I think we have demonstrated that we can be extremely profitable for a long period of time. Now coming to growth, in TMT we had a couple of client-specific issues because of which there was a decline in TMT. If we exclude TMT, our overall growth would be 27% for the year, so pretty good numbers overall. Within TMT, we had two client-specific issues. One of the clients did a joint venture because of which they stopped working with us or reduced the size dramatically, and the other one was a client that went through some reorganization and restructuring. So if you exclude those, overall growth numbers have been pretty good. Even including them, growth numbers are pretty good at about 19%. So that's what we think is pretty good for where Fractal is going from here.
R
Reema2:02
And those two TMT clients for which you're facing particular issues which you highlighted, does that persist in FY27 and will there be a quantifiable revenue loss on account of that?
S
Srikanth Velamakanni2:13
Those have now washed through the P&L. We don't see any more issues from those. So everything else comes back to normal. So we expect that that will have a good impact on the overall growth numbers for this fiscal.
R
Reema2:26
So FY27 revenue growth should be better than FY26.
S
Srikanth Velamakanni2:31
Better than FY26.
R
Reema2:35
Okay. You're not putting a number to it. Srikanth, morning. Just a quick word on, so two things. One, you were investing about 6.5% of revenues on R&D. That's 100 basis points more than what you did a year before. Where will this number settle at? And when we say R&D in the context of a company like yours, I mean where exactly is this going?
S
Srikanth Velamakanni3:02
Yes, let me explain that. Number one is that we expect to continue investing in R&D and increase the overall investments in R&D beyond where we are investing right now. That is the expectation. We want to take it up to 10% of revenue as R&D investments, which will be roughly between, let's say, companies that are services versus product companies which are at 14% of revenue spent on R&D. So we expect that number to go up. We want to do that while expanding gross margins. So as we expand gross margins, some of the gross margins we want to spend on increasing R&D spends. So the overall profitability continues to expand as we go forward. And in terms of where we are spending this money on R&D, we're building foundation models, agentic models. We're building this platform called Cogentic which is underneath every transformation solution that we're taking to. This is a multi-agentic platform where agents can work with each other, collaborate to solve a complex business problem. In doing so, they can access tools like machine learning algorithms and so on. And these algorithms in turn work with enterprise data. So this layer of agenticity on top of data platforms is critical to driving AI transformation inside companies. So we're building that platform. Secondly, we're building foundation models like our Vaidya medical foundation model which is doing really well on the HealthBench hard metric as well as the Asper platform which drives revenue growth for CPG companies. These are places where our R&D money is getting spent.
R
Reema4:39
Right, got that. And I'm assuming this is mostly, this 6.5% to 10%, right, you take the R&D expenses to, this is mostly on personnel, I'm assuming a large portion of it.
S
Srikanth Velamakanni4:55
Part of this is people and part of this is compute spends. So compute spends and including some spends on data acquisition and data labeling and things like that which are also spends. So overall it is about data acquisition, data labeling. It's a platform build which includes people as well as compute. And then certainly there is an R&D team that is also part of the expense overall.
R
Reema5:20
We are talking to you on a day, Srikanth, when this OpenAI deployment company is causing a bit of a flutter and IT services talks. You know, I was saying earlier maybe this was in response to Anthropic's lead on the enterprise side, maybe they want to show that they're doing enough as well, but of course it's having an impact here. Just your thoughts on this.
S
Srikanth Velamakanni5:42
This is, I think, the day Anthropic announced, OpenAI also announced the deployment company almost at the same time. So this has been a week in the making. It's a very good proof point that enterprises need a lot of work in order to drive value from AI. So what OpenAI and Anthropic both are saying is that our models alone are not enough. We need an AI services arm that can build this out for big enterprises and therefore they're investing in it because they don't see enough supply out there. So it's an enormous validation of the work that Fractal is doing, which means that this is a space that will take off quite well and with that Fractal as well. It almost feels like the model wars between Anthropic and OpenAI are behind us and now it's going to move on to implementation wars.
R
Reema6:31
But Srikanth, just getting back to the numbers then, if you want to increase your R&D spend to 10%, you will have to push your gross margins higher because that's where the money for investment into R&D will come. What will gross margins need to look like to make that 10% R&D spend? And secondly, even in this quarter we've seen an expansion in gross margins. Can you tell us how much of it came on account of pricing push towards outcome-based? And third lever is on internal productivity improvements.
S
Srikanth Velamakanni7:02
Most of this has come from moving from input-based models to outcome, output-driven models which are 5 to 7 points higher margin, and all license revenues are even 25-30 points higher in terms of margin. As we change the mix of business from input to output, all of the gross margin leverage is coming from that. Very little is coming from pricing-related advantages and almost nothing is coming from other aspects that you just mentioned. So overall it is about moving from input-driven models to output-driven models and we expect to continue that. And so we expect to grow faster, expand margins, and use some of the expanded margins into spending on R&D.
R
Reema7:41
So if I look at your current revenue base, what percentage of it would be input-based? Because that's the headroom you have to make that switch.
S
Srikanth Velamakanni7:50
Today about 60% of our revenues come from input-based models and 40% come from output, outcome, license revenue models. This 60 will drop to 40 or even under that, and that's the transformation that we are seeing. It could go down even further as well.
R
Reema8:08
And by when do you think?
S
Srikanth Velamakanni8:12
We have a three-year horizon for this. So the exact timing we are not dictating that because again we are a very client-centric organization. So we'll do what's right by the clients. But in general our shift is to propose an outcome or output-driven model as a starting point with every client.
R
Reema8:29
And pricing in the process of this shift, how will that change? So the pricing basis changes from input to output, but overall from a realization standpoint is per Fractal.
S
Srikanth Velamakanni8:42
Okay. And when you get to 40%, I mean the switch becomes 40% input, 60% output, what can gross margins look like?
R
Reema8:50
That extra 20% that shifts by about 7 points in terms of margin. So that is about 2 to 3 points of overall margin improvement from there.
S
Srikanth Velamakanni9:01
2 to 3% gross margin improvement.
R
Reema9:04
Yes. Got that. Two to three points. So 48 can become 51 basically.
S
Srikanth Velamakanni9:09
Got that. And this revenue per billable FTE, that's also slowly inching up 5%. Now with AI becoming so much smarter, do you think you can push revenue per billable employee say to 10% increase on a year-on-year basis?
R
Reema9:25
This is a very good question and one of the best metrics to look at any company today is are they expanding revenue per employee across sectors. It means that they are benefiting from AI and not just getting consumed by AI. So we would expect that the revenue per employee metric at Fractal should continue to go up substantially. Right now it's 5% but we expect that that number should be higher in the coming few years.
S
Srikanth Velamakanni9:49
All right Srikanth, we'll leave it there. Thank you very much for joining us. Good speaking with you and wish you the best of luck as you continue to execute.
R
Reema9:57
Thank you.