Fluid AI sells conversational and agentic AI software to some of the biggest banks on earth, including Barclays, Emirates NBD and, by the company’s own account, Bank of America — yet its founders say they have never taken a rupee of venture money to build it. That is the contradiction at the centre of this deep dive: an enterprise AI company competing for the same accounts as venture-fuelled rivals, while insisting it has been “profitable from day one.”
Brothers Abhinav and Raghav Aggarwal built Fluid AI out of a college-era learning platform, won a New York hackathon in 2012, and turned down at least two chances to take outside capital along the way. What follows is what could be verified about how a bootstrapped, Mumbai-headquartered AI vendor has stayed in business with global banking clients for well over a decade, and where the public record on it runs thin.
Quick facts
| Company | Fluid AI (built by Fluid Analytics; brands as Fluid AI) |
| Founded | Sources disagree: 2008–2009 as TruTech (Forbes company profile; Tracxn), relaunched as Fluid AI in 2012 after a New York hackathon win (Forbes India, 2017); the company’s own site header currently reads “Founded 2017” |
| Founder(s) | Abhinav Aggarwal (CEO) and Raghav Aggarwal (Managing Director) — brothers |
| Businesses | Enterprise “agentic” AI platform: conversational AI agents, workflow automation and AI-led analytics for banking, insurance, telecom and manufacturing clients |
| Latest FY revenue | Estimated under ₹10 crore (under $1.04 million) for FY25 (year to 31 March 2025), per third-party tracker Tracxn — not a company-published or audited figure |
| Latest FY profit/loss | Not publicly disclosed. Founders describe the company as profitable since inception (Abhinav Aggarwal, as reported by Forbes India, 2017) |
| Listed | Private — no stock exchange listing found |
| Market value / last valuation | Not independently verified this session; no confirmed valuation figure could be sourced from a page actually opened and read |
| Key shareholders / CEO | Abhinav Aggarwal (CEO); no external institutional investor identified in company records reviewed |
What they do
Fluid AI builds software that lets large, regulated enterprises — mostly banks — deploy AI agents that handle customer and employee interactions across chat, voice, WhatsApp and email, and that plug into legacy systems such as core banking platforms, CRMs and ERPs. Its own materials describe the current product as an “enterprise-grade agentic AI platform” for automating multi-step workflows rather than just answering questions, a positioning it has held since at least 2023–2024 conference appearances by its founders and now describes on its website (Fluid AI, 2026). The buyer is typically a bank’s or large corporate’s technology or customer-experience function, not a consumer.
The origin
The founding insight predates “Fluid AI” as a name. Abhinav and Raghav Aggarwal started their first venture, TruTech, in 2008 — a cloud-based learning management system built while Abhinav was still in school and Raghav in college (Forbes India, 2017). That business found an early market: by the mid-2010s it counted around 300,000 students using it, with a reported 95% retention rate (Forbes India, 2017). The brothers kept building on the side while both went on to enrol in MBA programmes at IIM Ahmedabad and ISB — only to drop out within a month of starting, once the AI business demanded full attention (Forbes India, 2017).
The pivot from an ed-tech tool to an enterprise AI company came through a hackathon, not a business plan. In 2012, the brothers entered a TechCrunch Disrupt hackathon in New York and won after a 24-hour coding sprint, and used that as the springboard to launch Fluid AI as a dedicated product under the TruTech umbrella, aimed at banks and retailers rather than students (Forbes India, 2017).
The struggle years
Selling artificial intelligence to banks in the early-to-mid 2010s was slow, unglamorous work. Fluid AI’s own account of that period, given to Forbes India in 2017, is that early sales cycles with financial institutions ran as long as six months per deal before the company managed to compress that to roughly one month — a change attributed to better-understood buyer objections and a sharper pitch, not to a single fix (Forbes India, 2017). There was no dramatic near-death funding crunch on record, because the company says it never depended on outside funding to survive in the first place; the struggle instead was the grind of closing enterprise contracts with no institutional runway to fall back on.
That constraint produced its own pressure-test moments. In October 2016, the brothers appeared on the Indian television pitch show “The Vault” and walked away from an on-air funding offer of ₹15 crore, judging the terms unfavourable to a business they had built without dilution (Forbes India, 2017). Around the same period, the company also fielded acquisition interest valued at roughly $100 million, and turned it down (Forbes India, 2017) — a decision covered further below. Both episodes are reported second-hand through founder interviews rather than independently confirmed deal documents, so they are presented here as founder-stated, not audited, history.
The turning point
If there is a single hinge moment in the public record, it is the 2012 hackathon win in New York. Before it, Fluid AI did not exist as a distinct product — the Aggarwal brothers were running a learning-management business serving students. After it, they had both a validated piece of technology and the confidence to build a second, separate company aimed at enterprise banking clients, launched under the TruTech parent (Forbes India, 2017). Within roughly five years of that pivot, the same 2017 profile reported the renamed company had grown to around 50 people, was working with banks on three continents, and was posting year-on-year revenue growth of about 300% for the fiscal year covered, with 250% growth projected for the following year — figures that, notably, come from the founders rather than a filed statement, so they are marked here as company-stated (Forbes India, 2017).
The money behind it
This is the section where Fluid AI departs most sharply from the typical “Deep Dive” subject: there is no ladder of priced funding rounds to walk through.
- No disclosed equity funding round: across the trackers checked this session, no priced venture round, lead investor or post-money valuation could be confirmed for Fluid AI (Tracxn profile, accessed 2026).
- One recorded grant: Tracxn’s funding history for the company lists a single entry — a grant dated 17 December 2020 from Microsoft Accelerator, with the amount undisclosed (Tracxn, 2026).
- Declined acquisition interest, reported value ~$100 million: Abhinav Aggarwal told Forbes India in 2017 the company had received acquisition offers valued at roughly $100 million and turned them down to stay independent (Forbes India, 2017).
- Declined on-air funding offer, ₹15 crore: on the television show “The Vault” in October 2016, the brothers rejected a ₹15 crore funding offer (Forbes India, 2017).
- Self-funded growth, by founders’ own account: “We’ve been profitable from day one,” Abhinav Aggarwal has said, with Raghav Aggarwal adding that the company ploughs back profits rather than raise outside capital, despite regularly fielding investor interest (as reported via Mastercard Newsroom, 2021).
No named venture-capital backers could be verified for Fluid AI in the sources opened this session. That absence is treated here as a data point about the company’s model, not a gap to be papered over.
How it makes money
Fluid AI is a business-to-business software vendor selling to large regulated enterprises, structured around a handful of mechanics that recur across the sources reviewed:
- Enterprise licensing, not consumer revenue: the customer is a bank, insurer, telecom operator or large corporate, not an individual end user (Forbes India, 2017; Fluid AI, 2026).
- Two historical product lines: an “experience” layer — smart-screen kiosks and interactive agents for customer-facing touchpoints — and a “data” layer, analytics software that scores customer behaviour for banks and retailers, per the company’s 2017 description (Forbes India, 2017).
- Current framing as an “agentic” platform: the company now describes its product as letting enterprises build autonomous AI agents that complete multi-step workflows — not just respond to queries — across voice, chat, WhatsApp and email, integrated into core banking, CRM and legacy systems (Fluid AI, 2026).
- Long, high-touch sales cycles: deals with banks reportedly still take roughly a month to close today, down from about six months in the company’s earlier years — implying revenue realised in discrete enterprise contracts rather than fast self-serve sign-ups (Forbes India, 2017).
- The part people get wrong: because its clients include marquee global banks, it is easy to assume Fluid AI operates at the scale and funding level of a venture-backed “AI unicorn.” The revenue and headcount data available point to a considerably smaller, founder-run operation than the client list alone would suggest.
No take rate, per-seat price or contract value could be sourced publicly for any Fluid AI deal, so none is quoted here.
The numbers
Fluid AI does not file the kind of disclosures a listed company would, and as a self-funded private company it has not published audited multi-year financials. The figures below are the only dated, sourced data points found this session — presented as-is rather than smoothed into an invented four-year run.
| Period | Revenue | Profit / loss | Source |
|---|---|---|---|
| FY17 (year to March 2017) | Absolute figure not disclosed; ~300% year-on-year growth reported, with ~250% growth projected for FY18 | Not disclosed | Forbes India, 2017 (founder-stated) |
| As of April 2021 | ~$1.7 million annualised revenue, on a team of 31 | Not disclosed; founders describe the company as profitable overall | Latka, 2021; Mastercard Newsroom, 2021 |
| FY25 (year to 31 March 2025) | Estimated under ₹10 crore (under $1.04 million) | Not disclosed | Tracxn, accessed 2026 (third-party estimate) |
No official profit-and-loss figure for any year could be verified from a source opened this session; every profitability claim above traces back to founder statements rather than filed accounts, and is labelled accordingly.
Where the money comes from
Fluid AI has never published a formal revenue split by geography or industry. What can be reconstructed is a client list, which skews heavily toward one sector:
- Banking and financial services (dominant, by named-client count): Barclays, Emirates NBD, Axis Bank and RBL Bank were named as clients in 2017 (Forbes India, 2017); the company’s current site adds HSBC, Bank of America and Mastercard to its client references (Fluid AI, 2026 — company-stated).
- Telecom and consumer corporates: Vodafone India and Intel were named as 2017-era clients (Forbes India, 2017).
- Professional services and industrials, referenced more recently: Accenture, Capgemini, Deloitte, Rolls-Royce, Toyota and Mercedes-Benz appear in the company’s current marketing materials (Fluid AI, 2026 — company-stated, not independently confirmed by those firms).
- Geography, by office footprint: headquartered in Mumbai, with additional offices reported in Charlotte, North Carolina; the Netherlands; and Mauritius as of 2017 (Forbes India, 2017).
The surprise is less about the split than about its staleness: the most detailed, named account of Fluid AI’s client base and office network is nearly a decade old. A company that names Bank of America and HSBC as references today has published almost nothing independently verifiable about its business since that 2017 profile — a pattern consistent with a firm that answers to founders and clients rather than to investors or public markets.
The risks
- A crowded, better-funded category: Tracxn’s competitive tracking places Fluid AI 403rd among 4,065 active companies in its category as of its most recent snapshot, naming Dialpad, Mistral AI and ClickUp as its closest tracked competitors (Tracxn, accessed 2026) — several of which have raised venture capital at a scale Fluid AI has not disclosed for itself, which can translate into faster product cycles and larger sales teams than a bootstrapped rival can match.
- Concentration in banking clients with long sales cycles: the named client list is dominated by large banks (Barclays, Emirates NBD, Axis, RBL, HSBC, Bank of America), and the company’s own account describes enterprise sales cycles of roughly one month to six months per deal (Forbes India, 2017) — a small number of large, slow-moving accounts leaves revenue exposed if even one or two contracts are not renewed.
- Capital discipline as a growth constraint: having declined a reported ~$100 million acquisition offer and a ₹15 crore funding offer to stay independent (Forbes India, 2017), and with only a single undisclosed-amount grant on record since (Tracxn, 2026), Fluid AI has less visible capital to deploy against product and go-to-market than venture-funded competitors in the same “agentic AI” category — a trade-off the founders appear to have made deliberately in exchange for control.
The takeaway
The lesson in Fluid AI’s public record is not about a dramatic rescue or a headline funding round — it is about what a founder gives up, and what they keep, by refusing outside capital in a market where nearly every competitor is raising it. Staying self-funded let the Aggarwal brothers turn down a nine-figure acquisition offer and a television funding deal on their own terms, and it forced the discipline of being profitable rather than merely growing. The cost, visible in a decade-old client list and a headcount still in the tens rather than the hundreds, is that the same discipline caps how fast the business can move in a category where rivals are spending venture money to grow faster than revenue alone would allow. Selling to banks does not require a startup to look, fund itself, or grow like one.
Frequently asked questions
Who founded Fluid AI?
Brothers Abhinav Aggarwal and Raghav Aggarwal founded the business, growing it out of an earlier venture called TruTech that they started in 2008; the Fluid AI product was launched in 2012 after they won a TechCrunch Disrupt hackathon in New York (Forbes India, 2017).
What does Fluid AI sell?
Enterprise software that lets large organisations, mostly banks, deploy AI agents for customer and employee interactions across chat, voice, WhatsApp and email, integrated with core banking, CRM and legacy systems (Fluid AI, 2026).
Has Fluid AI raised venture capital?
No priced equity round could be verified. The founders say the company has been self-funded and profitable since inception, and the only funding-related record found was a December 2020 grant from Microsoft Accelerator, amount undisclosed (Mastercard Newsroom, 2021; Tracxn, 2026).
Who are Fluid AI’s customers?
Banks and large corporates named across company and press sources include Barclays, Emirates NBD, Axis Bank, RBL Bank, Vodafone India and Intel (as of 2017), with the company’s current materials also naming HSBC, Bank of America and Mastercard (Forbes India, 2017; Fluid AI, 2026).
Is Fluid AI a public company, and what is it worth?
No. Fluid AI is privately held with no stock exchange listing found, and no valuation figure could be independently verified from a source opened this session.
Sources
Figures are as of September 2026. Currency converted at $1 ≈ ₹96.0 as of 18 September 2026 (Trading Economics).
- Forbes India, “The Future Is AI-ready,” Startups Special 2017 (forbesindia.com)
- Mastercard Newsroom, “AI tools for banks are growing. These two brothers in India are leading the trend,” 2021 (mastercard.com)
- Fluid AI, company “About” page, accessed September 2026 (fluid.ai/about)
- Tracxn, Fluid AI company profile, accessed September 2026 (tracxn.com)
- Latka, “How Fluid AI hit $1.7M revenue with a 31 person team in 2021,” accessed September 2026 (getlatka.com)
- Forbes, Fluid AI company profile, accessed September 2026 (forbes.com/profile/fluid-ai)
- Forbes, “30 Under 30 Asia 2017: Enterprise Tech,” April 2017 (forbes.com/pictures)
Found an error? Write to us and we’ll correct it in the open, dated, on the piece.