Last month, Fractal filed for a $2.5B Indian IPO to become the first AI company to list in India.

Genesis of the Algorithm

In 2000, the definition of “tech” in India was clear.

The best minds went into IT services. Infosys and Wipro were stock market darlings. Young engineers bragged about going on-site in the US. Families measured success in terms of H-1 B visas and dollar salaries.

Analytics? Nobody even used the word.

In 2001, five IIM-Ahmedabad classmates, Srikanth Velamakanni, Pranay Agrawal, Nirmal Palaparthi, Pradeep Suryanarayan, and Ramakrishna Reddy, pooled ₹2 lakh each and rented a one-bedroom flat in Andheri to start a consumer dot-com company, which was the fashionable play of the moment. Named after a mathematical phenomenon in which each part is the same as the whole, it was a good start for a technology company. 

But within months, the dot-com bubble had burst. Capital dried up, investor interest vanished, and the business plan collapsed before it could even take off. For most, that would have been the end.

Instead, the team made a sharp pivot. If consumer internet were not ready, they would bet on something almost nobody else in India was thinking about. Analytics to help enterprises make decisions based on data, not intuition.

It was a contrarian move. In 2000, banks still relied on branch managers’ instincts to decide who got loans. Insurance risk was assessed on paper files. Business decisions were “experience-driven”. 

Betting your career on analytics felt academic. But for the founders, that was where the opportunity lay. The first real validation was crucial because decisions had to be smarter if they were to get bigger and riskier. 

In 2001, ICICI Bank called Fractal in to build something no Indian bank had tried before, “a statistical credit risk scorecard”.

Until then, consumer lending decisions were made the old way. A branch manager would look at your file, maybe meet you in person, and then use his instinct to decide whether you were “good for it”. It was slow, subjective, and often incorrect.

Fractal’s model changed that. Instead of gut, it used probability. It analysed patterns in customer data and predicted the likelihood of defaulting. This was India’s first statistical model for consumer credit.

Once it worked for ICICI, others paid attention. HDFC signed them up. Even a Singapore bank, with access to much larger consulting firms, picked Fractal instead. These weren’t big projects in terms of revenue, they involved small teams and modest contracts. Strategically, however, they were huge. They proved two things. One, that decisions could actually move from gut to data, and two, that a tiny Indian startup could compete with global incumbents and win. 

Even then, with prominent brand names and logos, survival was a challenge. There was no cushion or a VC backer. By taking on "challenge projects" where clients only paid if they delivered results, the founders kept the lights on.

Imagine the pressure. Payroll depends on whether your model actually worked in the real world. As the founders pressed on, the lifeline came from outside the industry.

Gulu Mirchandani, from Onida, the same company known for its iconic devil ads, believed in their potential and wrote a cheque for $600,000. That money gave Fractal time. By 2005, revenue had finally surpassed $1M. Still tiny compared to IT services, the revenue was validation that this contrarian bet could compound, especially in a world where “analytics” wasn’t even a proper category yet.

But another truth had become evident by then. Indian clients were curious but not generous. They loved pilots and presentations, but when it came to aligning on really big money, they hesitated. In the US, it was different. If you showed results, budgets followed.

In 2005, the founders shifted Fractal’s commercial base to New York while maintaining delivery operations in Mumbai. 

It meant Srikanth and Pranay were suddenly pitching Fortune 500 clients face-to-face, not over long-distance calls. It meant that in the eyes of global clients, Fractal was no longer a scrappy Andheri experiment but a company with a US footprint.

Decoding the Noise

By 2007, Fractal had grown to 100 employees, generating approximately $4M in revenue. 

No longer a side experiment, it was now a real company. That’s when the first rupture came. As the scale grew, leadership could no longer run on consensus. The founders held a process to pick a CEO. Srikanth Velamakanni was chosen. Soon after, two co-founders, R.K. Reddy and Pradeep Suryanarayan, left. The exits were painful. Some team members followed. 

However, the split necessitated discipline in governance, accountability, and structure. What began as a friendship-driven initiative was now evolving into an organisation.

In 2006, Procter & Gamble floated an RFP for marketing-mix modelling. Over 25 firms applied. Fractal, still young, won a six-month pilot. Their models optimised media budgets by 35%, driving revenue growth of 5% or more. By 2008, P&G had integrated Fractal into its futuristic Business Sphere dashboards, which served as the central hub for making billion-dollar marketing decisions.

An Indian startup collaborated with one of the largest CPG firms globally. Symbolically, it was a breakthrough. Analytics from an India upstart competing head-to-head inside a Fortune 500 boardroom.

All would not be smooth sailing, as startups are generally not linear. 2009 would be a body blow.

The global financial crisis hit, and General Motors, one of Fractal’s biggest clients, filed for bankruptcy. Almost 10% of revenue vanished overnight.

Survival instincts took over. The company pivoted to credit-recovery models for US auto lenders, helping banks manage defaults as households collapsed under debt. That agility kept them alive.

By 2010, growth returned to ~35% year-over-year. But the scars remained. Losing a major client overnight prompted Fractal to reassess its business approach. They learnt to avoid overdependence, maintain cash reserves, and always hedge their risks.

The first decade didn’t make Fractal a big company. But it made them durable. From India’s first credit models to a seat at P&G’s decision rooms, from founder exits to recession scars, they built a spine. 

That spine became the foundation for everything that followed.

Seed Data, First Signal

By 2010, Fractal had a decade of scars to show. 

It had survived the dot-com bust, the departure of two founders, and the near-death shock of losing General Motors in the financial crisis. What it lacked was scale capital.

The founders had begun serious fundraising in 2006–07, approaching marquee venture firms. The meetings were promising at first. Investors acknowledged that analytics could be a sector of the future. But every term sheet carried the same caveat: “resolve the founders’ dispute first.” 

Some in the founding team wanted Fractal to grow conservatively in India, milking small but steady contracts. Others argued for blitzscaling abroad, burning cash to win global share. Neither side yielded. By the time the dispute ended, the 2008 crisis had frozen capital markets. The window was gone.

That left only one option, to grow profitably. 

Easier said than done. Clients admired Fractal’s work but viewed it as a boutique vendor rather than a partner. One investor crystallised the perception “You are the most unfriendly company we have ever met.” The remark forced reflection.

Change began with a coach. The founders brought in Shreekant Gupte, a former Marico executive with scars of his own from running divisions in competitive FMCG markets. He reframed Fractal’s approach through what became the “Total Client Solutions” strategy. Instead of spreading thin across new logos, Fractal would go deep. 

The idea was simple. Serve a handful of clients so well that you became impossible to replace. Stop being the lowest-cost analyst and start being a strategic partner.

Inside the company, culture became the next battlefield. Fractal codified its People Principles, which included trust, transparency, and autonomy. In an industry where analysts were often treated as billable units, Fractal wanted to signal that talent was a growth partner, not a resource. 

It was a gamble that cultural stickiness could translate into client stickiness.

The results were slow but steady. Between 2011 and 2013, Fractal focused almost entirely on expanding within existing clients. The bet paid off. Revenue inched up to $8M by 2013, still modest but with healthier margins and deeper relationships.

That year, the clouds finally parted. 

TA Associates, one of the world’s leading private equity investors, wrote a $25M investment. The money gave Fractal firepower to expand teams, invest in technology, and strengthen its US and UK presence. Symbolically, it was a turning point. 

After years of being seen as a scrappy survivor, Fractal was now recognised as the company leading India’s charge in analytics.

The Billion Dollar Dataset

By 2016, the world of data, AI, and analytics had already paved the way for enterprise transformations. 

AI was entering its newest wave of excitement. Industry reports pegged the global AI and analytics market at over $50 billion in the mid-2010s, with projections soaring to $3 trillion by 2024.

Fractal, now a pioneer in analytics with a proven track record, has demonstrated its ability to translate insights into impact by combining deep analytics expertise.

But the opportunity ahead required more than services. Not only did the company have to identify the problem and solve it with a consulting mindset, but it also had to build solutions that could scale.

Effectively, Fractal needed to evolve from a consulting-led analytics firm into a full-stack AI powerhouse. In May 2016, Fractal achieved a significant milestone when it secured a $100M investment from Khazanah Nasional, Malaysia's sovereign wealth fund.

The investment gave Fractal the firepower to double down on R&D, expand its deep learning capabilities, and accelerate an inorganic growth strategy that would redefine its trajectory.

The funds set off a chain of bold moves.

Fractal reorganised its business in a structure designed to scale solutions while staying close to client problems. It incubated Qure.ai to revolutionise healthcare diagnostics with AI and launched Cuddle.ai to simplify executive decision-making. 

Products like Samya and Crux followed, reflecting a clear ambition: to own the decision intelligence layer for enterprises worldwide.

What began as an analytics services firm was now morphing into a company that could build, consult, and deliver end-to-end AI ecosystems. 

Models and Margins

By 2017, Fractal had moved past the phase of proving itself. 

It was no longer just another analytics outsourcer working on project contracts. The company had begun to lock in multi-year partnerships with Fortune 500 firms. These contracts gave it stability, the kind of dependable inflow that allowed it to think bigger.

Bigger meant reimagining its economics.

Fractal had stumbled upon a hybrid model that looked deceptively simple but was rare in India’s analytics space. Consulting projects became the base. These engagements with consumer goods giants, banks and insurers ensured recurring revenue.

But instead of just scaling billable hours, Fractal started to funnel nearly 12–15 per cent of its top line back into research. This was a bold allocation. Peers like LatentView or Tredence spent less than half that amount, and global firms like Mu Sigma preferred to lean on service delivery rather than internal IP.

The bet was on spin-outs. It had already started with Qure.ai in medical imaging, Cuddle.ai in business intelligence, and Trial Run in marketing simulation, which were not side projects.

They were structured to become high-margin products, capable of scaling without being tied to headcount. Consulting generated cash flow, and the products promised operating leverage.

Unit economics during this phase began to reflect the shift in strategy. The revenues grew from roughly 420 crore in 2017 to over 800 crore in 2019, and margins moved in tandem. 

For a services-heavy firm, this was a crucial turning point. Employee costs still took up nearly seventy per cent of revenues, but revenue per employee was climbing steadily as global delivery matured. What might have been close to 20 lakh per head in 2017 was edging towards 25 lakh by 2018, showing better productivity and utilisation.

This was also the time Fractal began sharpening its offerings by acquiring companies at its edges. In 2017, it acquired Chicago-based 4i Inc., a growth strategy and analytics firm that added depth to its consulting capabilities. 

A year later, it brought in Final Mile, a Mumbai- and US-based firm steeped in behaviour science. That acquisition was unusual for the analytics industry. Final Mile’s expertise in cognitive neuroscience and behavioural economics meant Fractal could not only crunch numbers but also understand and shape human decisions. For a global enterprise, that combination was potent.

Multiple consecutive quarters of growth through 2018 demonstrated that the strategy was working. Fractal was now seen as more than a vendor. It was positioning itself as a partner that could deliver outcomes by combining data, algorithms, and behavioural design.

Consulting was the safety net. Intellectual property was the moonshot. Together, they began to harden Fractal’s margins and clarify its business model.

Scaling the Neural Net

While the previous years focused on fine-tuning the engine, 2019 marked a significant acceleration for Fractal.

The company received a capital infusion of approximately $200M in 2019, valuing it at $500M. This was not just liquidity for the early backers but fresh firepower for the business itself. 

The mandate was clear. Build technology, expand capabilities and chase global scale.

Revenue climbed from around 800 crore in 2019 to 1,300 crore by 2021. That is a compounded growth rate of nearly 30 per cent at a time when the global economy was wobbling. Fractal was growing fast and profitably.

Acquisitions became a tool of precision.

In 2021, Fractal acquired Samya.ai, a company specialising in revenue growth management for consumer goods. The problem it solved was massive: CPG firms were losing billions of dollars annually to ineffective promotions and demand forecasting. Samya’s deep learning models promised to claw that back. 

Later that year, Fractal invested in Senseforth.ai, a conversational AI platform that automated millions of customer interactions for enterprises. Although the deals were not significant in terms of revenue, they integrated new capabilities into Fractal's existing portfolio.

These sat on top of an internal portfolio that was already rich. 

By 2021, the Ideas2Business incubator had spun out multiple products that it had identified as opportunities five years prior. Qure.ai for healthcare imaging, Theremin.ai for financial alpha, Cuddle.ai as a business co-pilot, and Eugenie.ai for anomaly detection.

With Samya and Senseforth joining in 2021, Fractal could now address everything from demand forecasting to customer engagement. What began as consulting projects had evolved into a full suite of AI-driven solutions, each designed to solve specific pain points inside the world’s largest corporations.

This expansion went hand in hand with scale.

By now, Fractal had thousands of employees across the US, UK, India, Singapore and Australia. Nearly 70 per cent of revenues came from the US, where clients such as Wells Fargo, P&G, and Visa continued to expand their contracts. 

Its economics underpinned the global delivery model. Efficient delivery centres in India balanced premium client-facing work in the Americas. As a result, revenue per employee was estimated at close to 30 lakh by 2021, well ahead of Indian peers and approaching global benchmarks.

Employee expenses were still the largest line item, but with multi-year client relationships averaging nearly a decade, customer lifetime value was high and payback periods short.

This was also the time when AI stopped being a buzzword and started running at scale. 

Banks deployed conversational bots built with Senseforth to serve millions of retail customers. Consumer goods giants utilised Samya’s deep learning to optimise promotions and recover value that had long been lost through discounts and stockouts.

Healthcare providers began adopting Qure.ai’s imaging platform for diagnostics, demonstrating that AI can operate effectively within regulated, high-stakes environments.

By 2021, Fractal had outgrown the tag of an analytics services firm. It was now a diversified provider of AI solutions. Its consulting muscle provided credibility, its IP portfolio offered scale, and its acquisitions sharpened its edge.

In just a few years, Fractal had gone from crunching data to architecting the future of enterprise decision-making. The neural net it was building was no longer just mathematical. 

It was organisational, commercial and global.

Training Against the Competition 

By 2022, artificial intelligence had become a priority for every large enterprise. Technology budgets were expanding, and companies were actively seeking partners who could help them adopt AI at scale. 

This also meant competition for Fractal intensified. 

It was no longer competing only with Indian peers, such as Mu Sigma. Global consulting firms such as Accenture and Deloitte have made analytics central to their digital transformation practices. Specialist players, such as Palantir, were pushing enterprise platforms, while new SaaS startups offered packaged AI products that promised quick deployment.

Fractal chose not to compete by breadth but by depth. It focused on verticals where it had built expertise over the years for consumer goods, healthcare, and financial services. 

Another differentiator was its decision-first approach. 

The insights alone were not enough; the challenge was ensuring organisations acted on them. The acquisition of Final Mile, a behavioural science firm, gave Fractal a unique capability to design interventions that influenced human decision-making inside companies. This positioned Fractal not just as a data analytics provider, but as a partner in driving measurable outcomes.

The company also invested heavily in intellectual property. 

By this time, it had filed more than 50 patents worldwide. It continued with a high level of reinvestment, which was unusual for a services-led company. It allowed Fractal to build proprietary frameworks and solutions that could not be easily replicated.

In January 2022, Fractal secured a $360M investment, valuing the company at over $1 billion. This funding made Fractal a unicorn, providing the financial strength to accelerate acquisitions, expand into new geographies, and scale its platforms. More importantly, it gave clients confidence that Fractal had the backing and resources to deliver at the same level as the world’s largest consulting and technology firms.

Financially, these years were still about striking a balance between growth and investment. 

In FY23, Fractal reported revenues of approximately ₹2,043 crore, with EBITDA margins of nearly 22%. The company, however, posted an operating loss of approximately ₹210 crore, as it continued to spend on R&D and acquisitions. The funding gave it the cushion to sustain this strategy, prioritising capability-building over short-term profit.

By FY24, Fractal was working with over 100 Fortune 500 clients. Its positioning was precise: a specialist AI partner with deep domain knowledge, strong IP, and a proven ability to embed AI into real decision-making processes. 

This clarity helped it stand apart in a market that was becoming increasingly competitive.

The Current Dashboard

By FY25, Fractal had shifted from being an AI thought leader to a fully scaled enterprise AI powerhouse. 

The company reported consolidated revenues of INR 2,765 Cr, growing at 26% year-on-year – its highest organic expansion since the pandemic. Profitability rebounded as well, with net income hitting INR 220 Cr, reversing a dip in FY24. At a time when many AI-native startups were chasing scale at the expense of stability, Fractal was growing both its top and bottom lines.

Fractal was now working with 177 enterprise clients, including over 100 Fortune 500 companies across the CPG, financial services, technology, and healthcare sectors. The US accounted for 65% of the revenue, while Europe accounted for 16%, and India and the Asia-Pacific region made up the rest. Importantly, repeat business made up over 90% of its revenues, with the top 10 clients contributing more than 30%.

A big reason behind this stickiness was Fractal’s reorientation toward verticalized solutions. Each industry, whether it be insurance, retail, or pharmaceuticals, had a tailored solution team, staffed with both domain specialists and AI consultants. That improved win rates and retention, especially as enterprises began demanding contextual, industry-specific AI rather than generic tooling.

On the product front, Fractal consolidated its IP into two major platforms. Fractal Alpha, its decision intelligence suite, and Cogentiq, a secure, agentic AI platform for enterprise assistants. Cogentiq gained significant traction in 2024, as enterprise clients sought conversational AI that could execute workflows, not just answer questions. These were not B2C chatbots. They were operational copilots designed for enterprise-grade complexity and compliance.

The foundation for these platforms had been laid years earlier. Now, products like Qure.ai, Samya.ai, and Senseforth.ai had matured. They were no longer moonshots. They were revenue contributors, embedded inside client workflows.

Internally, Fractal matured just as quickly. 

Headcount rose to 5K+, and 17% of employees owned ESOPs, reinforcing a culture of shared ownership. Its global delivery centres in India remained central to cost control, while solutioning and client teams in the US and Europe deepened customer intimacy. Revenue per employee crossed INR 55 lakhs, a sign of increased operating leverage.

By the end of 2024, Fractal was operating in a rare space.

The company was profitable, product-driven, and globally relevant. With expectations that the enterprise AI market will surpass USD 200 billion by 2026, Fractal was not merely following the trend. They were building the infrastructure to steer it.

Future State

In August 2025, Fractal filed its DRHP, aiming to raise INR 4,900 crore through a mix of a fresh issue and an offer-for-sale. If successful, the IPO would value Fractal at over 25,000 Cr, making it one of India’s first publicly listed enterprise AI companies.

Fractal’s north star is to become the operating system for enterprise decision-making. The company is pursuing a three-pronged growth strategy: deepening productisation, expanding globally, and making strategic acquisitions.

Product-wise, Fractal is going all-in on Fractal Alpha and Cogentiq. 

The former aims to be a modular platform that spans strategy, forecasting, and experimentation i.e., a full-stack decision infrastructure. The latter focuses on building AI agents embedded in enterprise functions - marketing ops, supply chain optimization, and finance automation.

These are not horizontal tools. They’re being built to operate within regulated, complex industries where explainability, integration, and ROI are key. This means investing heavily in R&D, cloud engineering, and domain-specific models. Over 15% of projected post-IPO spending is earmarked for product and research.

M&A will be the next lever. Fractal is already scouting for acquisitions in cloud AI, life sciences, and telecom analytics, verticals where client pain points are high, and AI penetration is still low. Smaller tuck-in acquisitions in Europe and Southeast Asia are also being explored to grow regional presence.

Geographically, Fractal aims to strengthen its onshore presence in Europe and the Middle East while expanding delivery hubs in Tier 2 Indian cities to reduce attrition and increase cost efficiency. By 2026, it expects to grow headcount to 7K+, with hiring focused on product, design, and AI engineering.

Strategically, Fractal does not want to become a SaaS-only firm. Nor does it like to be a pure consulting company. It aims to remain hybrid, combining consulting with platforms. The consulting arm earns trust and uncovers latent business needs. The platforms solve them at scale.

That is why, internally, the vision is often described as ‘the Accenture of AI.’ A long-term partner for CEOs looking to embed intelligence into every business decision, not just run AI pilots, but transform enterprise workflows end-to-end.

The company also plans to open-source select components of its decision stack to create a developer ecosystem around enterprise AI tooling. While this will take time, the goal is to ensure that Fractal not only delivers tools but also sets the standard for how decision intelligence is built and measured.

Fractal’s journey over 25 years has been anything but formulaic. It didn’t chase consumer hype or burn capital on moonshots. It stayed in the trenches, solving challenging enterprise problems, one workflow at a time.

It bet early on AI. It invested in IP when others optimised margins. It built products without abandoning consulting. And it scaled globally without losing profitability.

Now, with IPO capital, platform maturity, and industry tailwinds, Fractal is poised to take on its most significant role yet. To become the invisible yet essential layer powering decisions within the world’s most complex businesses.

The journey to get here has taken 25 years, from 5 college mates to a 25,000 Cr company. The path to building is never linear, but if it works, it can be huge. 

Fractal could become the company powering AI for enterprises, building from India to the world. 

Writing: Parth, Ajeet, Ananya, Nikhil, Tanish and Aviral Design: Omkar