Key Takeaways
- 26% 9M revenue growth, both segments even: Banking & Fintech and TIG each grew 26%, dispelling any notion of a “flattish” banking performance; quarter-to-quarter variation is normal.
- Deliberate capacity squeeze underway: Headcount is being held flat to force an AI-native reset of the software delivery model, creating near-term delivery pressure that may constrain revenue recognition until productivity gains arrive.
- AI investment cycle is concentrated and multi-vector: R&D stays at 9–10% of revenue, intangibles are building for AI and payments, and management is pursuing three ROI horizons: Banking Software 2.0, agentic process execution, and internal AI-native transformation.
- Transit is becoming a margin story, not just a growth story: ‘Make in India’ hardware capability, IP ownership across the value chain, and a growing software component are steadily improving transit margins, with management expecting them to at least match banking software margins over the long run.
- What to watch next quarter: Whether Q4 go-lives and project executions convert the large pipeline into cash and turn operating cash flow positive for the full year, and any signal on when the headcount-driven capacity constraint begins to ease.
The AI-Native Reset: Strategy, Investment & Capacity
- Headcount deliberately decoupled from revenue: Management intends to largely decouple headcount growth from revenue growth, driving operating leverage without materially expanding the workforce in most businesses over the next year. This is not a cost-cutting exercise: it is a forced reset to embed AI tools and new ways of delivering software across the organization.
- Short-term pain for productivity gain: The decision not to increase headcount has created delivery capacity pressure that may constrain near-term revenue recognition. Changing from a legacy software build process to an AI-native one takes time, and the alternative of simply adding headcount would remove the pressure to change.
- Core development headcount has actually come down: On the value-adding side, headcounts have shrunk; management expects productivity to enhance significantly once the reset takes hold, but the timeline is uncertain: “Is it tomorrow? Is it next month? Is it next quarter? I think time will tell.” At most a few quarters.
- Three ROI vectors with different horizons: (1) Banking Software 2.0, an AI-native stack with near-term ROI over 2–3 years, already visible in wins; (2) agentic execution of processes via AurionAi, a medium-term prize; (3) becoming an AI-native organization internally, the most immediate ROI over the next few quarters through productivity enhancement and lower cost of production.
- R&D spend held at 9–10% of revenue: Management is reimagining the R&D model and software supply chain by embedding AI-led intelligence across the entire software development lifecycle.
- Intangible asset build reflects AI and payments bets: Roughly ₹100 crore in combined H1 intangibles came from the AuroPay payments business product build-out (anticipating regulatory licenses) and significant AI investments across AurionAi, Orion-MSP, and Lexsi Labs research. No capitalization has been done around core product lines like transaction banking or lending.
- AI investment is a moving goalpost: While the payments-side build is “largely good,” AI investment will continue; management may seek government cross-subsidization but that takes time. No clear commitment on when or whether AI-related capitalization will stop.
Ashish Rai: “If you want to become an AI-native software firm, it is not just a question of starting to use Cloud, Code, or Cursor or whatever your favorite tool is, it is a question of how do I fundamentally reimagine the workflow?”
Banking & Fintech: Wins, Stickiness & the AI Overlay
- Both segments grew at the same 26% clip: Management pushed back on any characterization of banking growth as flattish, noting that both Banking & Fintech and TIG have grown 25–26% over the last three quarters. Banking revenue conversion cycles are far shorter than transit, so smaller-looking deal values still translate into growth.
- Landmark lending win in Singapore: A leading Singapore-headquartered bank chose Integro to modernize its lending operations across international subsidiaries, a mandate that validates the platform’s cross-border capability.
- iCashpro set up as the market standard: A marquee Indian public sector bank selected iCashpro as its preferred transaction-banking platform, a win management says “pretty much sets up iCashpro as the market standard” for new deals. CSB Bank also signed for a next-generation cash management solution.
- IDBI Bank data center project: A strategic data center win marked Aurionpro’s entry into IDBI Bank, adding a new logo in a relationship that could broaden over time.
- InfraRisk acquisition fills lending gaps: The Australia-based InfraRisk brings commercial and auto lending capability, strong clients including Judo Bank and Toyota, and a global team. It rounds out a full-spectrum lending play.
- Software stickiness is structural: Complex enterprise banking software is extremely hard to replace, and heavy regulation creates inertia. Management sees no new rationale to replace it in the medium term (4–5 years) unless legacy systems get in the way of AI adoption. The revenue stream is extremely sticky.
- AI enlarges the opportunity, it does not disrupt the incumbent: The playing field is “fundamentally tilted towards the incumbent software provider” in banking because the domain is so complex. AI creates a need to enhance software (Software 2.0), increasing the size of the opportunity rather than threatening it. Aurionpro has invested accordingly: the Arya.ai acquisition, labs in London and Paris, and a recent patent on DL Backtrace for AI explainability.
- Revenue model is built for recurrence: For a typical software deal, roughly one-third is license, one-third is AMC, and one-third is a 12-to-18-month project. Licenses are five-year term licenses, but banks buy for 20–30 years, so every year roughly 20% of the base comes up for renewal, making even term licenses recurring at a five-year frequency.
Ashish Rai: “I do not see anything changing at all in the medium-term, which is, let us say, four or five years. I see a need to enhance the software to do more. Hence, the Software 2.0 story.”
Transit: Deals, Margins & the ‘Make in India’ Edge
- TIG growth driven by large deals: Strong growth in the TIG segment has been fueled by large wins over the last couple of quarters, including MMRDA, Titagarh (platform screen doors for Mumbai Metro Line 5), and MP Metro (automated fare collection for Bhopal and Indore). The pipeline remains very large.
- Transit margins are steadily improving: Three structural drivers: (1) Aurionpro now owns IP on every point of the value chain it competes on; (2) ‘Make in India’ production capability on the hardware side improves economics; (3) a growing software component in deals (e.g., CCHS for MMRDA) provides operating leverage.
- Long-run margin ambition: Management expects transit to at least match banking software margins over the long run, and it could become even more profitable depending on the stable state.
- DSOs are manageable: Transit organizations have been good at turning around payments; DSOs are not expected to exceed current TIG or enterprise levels. Once in steady state on long-term contracts, money comes in predictably. The company is more careful now about what it signs, and DSOs have been tightening a few days every year.
- Revenue models include revenue share: Beyond licenses and subscriptions, some transit deals carry revenue-share structures. Model choice is centered on maximizing ROI for the investment made, not maximizing the next quarter’s dollar.
Pipeline, Data Centers & the Demand Environment
- Largest sales quarter in company history: Q3 produced the highest sales quarter on record, with strategic wins across both core segments and 23 new logos added, one of the fastest rates of customer additions ever.
- Pipeline is 65–66% bigger than last year: Management declined to give an absolute pipeline number, calling it “misleading,” but disclosed the year-on-year growth rate. The pipeline could be “several thousands of crore” but that figure is not useful to shareholders, per management.
- Data center demand far outstrips capacity: The demand environment is significant and unlike anything seen before. Aurionpro wants to be the go-to partner for highly complex, specialized design and build projects, not a commodity player. The company is looking to expand its number of strategic partners and expects to share more news in the coming months.
- Government budget push is a tailwind with uncertain scale: On a three-to-seven-year horizon, management expects a lot more investment in the market given the government’s budget push, but candidly admits “the honest truth is I do not know” whether it will grow the opportunity to a totally different scale.
- Data center business to grow faster than enterprise rates: The segment is expected to outpace broader enterprise growth over the next few years.
Ashish Rai: “The question really is not the size of the opportunity. The question is, what role does Aurionpro want to play? We want to be the go-to partner for highly complex, specialized design and build projects.”
Financials: Margin, Cash Flow & Capital Allocation
- 9M EBITDA up 23% to ₹216 crore: Margins held at 20% plus, reflecting continued operating leverage even as the company invests in AI and payments.
- Q3 PAT margin dipped on a one-off: PAT margin was 12% for the quarter versus 14% for the nine-month period, due to a one-off cost item from implementation of the new labor code.
- H1 operating cash flow was negative ₹80–86 crore: Management characterized this as typical mid-year seasonality, with more projects running in parallel this year than last. Q4 is a big quarter for go-lives and project executions; so far on track. The company is likely to end the year with positive operating cash flow, barring something dramatic on execution.
- Cash balance roughly ₹250–300 crore: Management did not have the exact number but provided a range. Other income of ₹2 crore looks low relative to the cash balance; management acknowledged it had not looked at this deeply and offered to return with specifics.
- Acquisition strategy scales with the company: Aurionpro will continue making acquisitions to the extent it has free cash to invest. Every net-new space involves a buy-versus-build choice. The scale of acquisitions will get bigger because the size of deals the company can now do is larger.
- No specific recurring revenue percentage disclosed: Management explained the revenue model instead of giving a number, emphasizing that term licenses are effectively recurring at a five-year frequency given banks’ 20–30-year relationships.
Competitive Position & Long-Term Ambition
- Five-year transformation from services to IP-led: Five years ago the company was services-heavy with 2–3 salespeople; today it has 45–46 salespeople, partnerships with MasterCard, Finastra, and Visa, and is doing deals of ₹250 crore scale that were unimaginable 12 months ago.
- Goal is top three globally in each space: Management’s stated ambition is to be a top three global player in each of the segments it operates in, while acknowledging “you are not even close.”
- AI expands the addressable market dramatically: Management estimates the transition toward AI-driven reasoning expands the relevant market from roughly $400 billion to $2.5–$3 trillion.
- Orion-MSP is a top-tier tabular foundation model: Lexsi Labs’ Orion-MSP ranks No. 1 or No. 2 globally on most benchmarks for tabular data, per management. It launched alongside TabTune, an open-source library for scalable enterprise AI on structured data.
- AuroPay licenses received in India: Both offline and online licenses are now in hand; progress continues in other geographies.
Ashish Rai: “We are still a tiny mosquito in a giant industry. We just have to be a fierce, we just have to be a competitive fierce mosquito, and we will continue to grow.”
Guidance & Commitments
| Commitment | Specifics | Timeline |
|---|---|---|
| Revenue growth trajectory | Grow at roughly 30% annual clip, with year-to-year variability (e.g., 34% one year, 26% another) | Over the next few years |
| Headcount decoupling | Largely decouple headcount growth from revenue growth, driving significant operating leverage without materially expanding workforce in most businesses | Over the next year |
| R&D spending | Remain in the range of 9% to 10% of revenues | Ongoing |
| Operating cash flow | End the year with positive operating cash flow, barring something dramatic on execution | FY26 full year |
| Transit margins | At least match banking software margins over the long run; could become more profitable | Long run |
| Data center growth | Continue growing faster than enterprise rates | Over the next few years |
| Productivity gains from AI-native reset | Productivity enhancement and cost of production to come down | At most a few quarters |
| Acquisition scale | Continue acquisitions; scale will get bigger as the company’s deal capacity grows | Ongoing |
Key Metrics
| Metric | Q3 FY26 |
|---|---|
| Revenue from operations (9M FY26) | ₹1,066 cr, +26% YoY |
| Revenue from operations (Q3) | ₹371 cr, +21% YoY |
| EBITDA (9M FY26) | ₹216 cr, +23% YoY; margins at 20%+ |
| EBITDA (Q3) | ₹75 cr, +18% YoY; margins steady at 20%+ |
| Profit after tax (9M FY26) | ₹150 cr, +9% YoY; PAT margin at 14% |
| Profit after tax (Q3) | ₹44 cr; PAT margin at 12%, reflecting a one-off cost from new labour code implementation |
| Banking and Fintech segment growth (9M FY26) | 26% YoY |
| Technology Innovation Group segment growth (9M FY26) | 26% YoY to ₹471 cr |
| R&D spending | 9% to 10% of revenues, expected range going forward |
| New logos added | 23 across business segments |
| Order book | ~₹1,650 cr or slightly more |
| Pipeline growth | 65% to 66% larger than same point last year |
| Cash balance | ₹250–300 cr range |
| Operating cash flow (full year expectation) | Positive, barring dramatic execution issues |
| DSOs | 100 to 120 days |