Startup Deep Dive : Entropik — revenue tripled to Rs 47.53 crore in FY23 even as losses widened to Rs 27 crore

In FY23, a Bengaluru company that reads faces, eyes, voices and brainwaves grew its revenue nearly three times, to ₹47.53 crore (about $5.0 million at $1 ≈ ₹96.0, 18 September 2026, Trading Economics) as per its filings compiled by thekredible. In the same year its net loss widened to ₹27.2 crore, as reported by Inc42 from the company’s MCA filings. That is the paradox at the heart of Entropik Technologies: the more it sold, the more it spent to sell it.

Entropik pitches something that sounds closer to neuroscience than to a research vendor: “Emotion AI” that infers what consumers feel from facial coding, eye tracking, voice tone and, in its earliest lab product, brainwaves. The founders began the idea as a student project at IIT Kharagpur and turned it into a SaaS platform used by brands such as KFC, Mondelez, Reckitt and P&G, per the company’s own client list. Investors including Bessemer Venture Partners and SIG bought the story to the tune of roughly $34 million across six rounds since 2016, per Inc42’s funding data. What follows is what the public record actually supports, and where it goes quiet.

Quick facts

What Entropik does

Entropik sells research software to the people who spend real money understanding customers: brand, insights and product teams. Instead of relying only on what respondents say, its platform reads how they react. According to the company, its Emotion AI stack combines several signals:

• Facial coding — reading micro-expressions from a webcam to infer emotional response to an ad, pack or screen.
• Eye tracking — mapping where attention lands and for how long, via a standard camera.
• Voice/tonal analysis — inferring sentiment from how respondents speak, not just the words.
• Brainwave (EEG) analysis in its earliest lab product, Affect Lab, for deeper neuro-research use cases.

That technology is packaged into products aimed at two buyers, per the company: Decode, an integrated consumer and market-research platform, and Qatalyst, a user-experience/UX research platform. The company lists brands including KFC, Tata AIG, Mondelez, Reckitt, Cipla, P&G, Disney and Kellogg’s among its users. Entropik describes itself as being among the top emotion-technology companies globally; that ranking is company-stated and not independently audited here.

The origin

The seed of Entropik was not a boardroom, but a campus. Ranjan Kumar traces the idea to a student project at IIT Kharagpur, where the question was whether technology could measure how people actually feel rather than what they claim to feel on a survey. He did not go straight from lecture hall to launch. Per his public profiles and company accounts, he worked across large organisations and payments before returning to the emotion idea, and had already tried his hand at earlier ventures. When he founded Entropik in 2016 with Lava Kumar, the pitch was narrow and specific: traditional market research asks people to rationalise decisions they made emotionally, and that gap was worth a business.

The early bet was hardware-adjacent and lab-heavy. In 2018 the company launched Affect Lab, which it positioned as a multimodal emotion-AI platform stitching together facial coding, eye tracking and brainwave data. It was a credible research tool, but it was also a slow, high-touch way to sell software. The origin insight was right; the packaging would take years, and two full product rewrites, to match how customers actually wanted to buy.

The struggle years

Entropik’s difficulty was never demand for the idea. It was turning a lab into a repeatable, self-serve product, and paying for that transition out of a small revenue base. The company raised a modest $200,000 seed in November 2016 and a $1.1 million round in July 2018 from Bharat Innovation Fund and IDFC-Parampara, per Inc42 and BW Disrupt. That is not much runway for a company trying to build computer-vision and signal-processing models from scratch.

The financials of the early scale-up show the strain plainly. Revenue from operations was ₹7.98 crore in FY21 against a net loss of ₹2.91 crore, as reported by Entrackr. As the company shifted from a services-flavoured lab model toward a software platform, spending ran far ahead of income: by FY22 the net loss had jumped roughly six times to ₹17.7 crore, per Entrackr, even though revenue was still only in the mid-teens of crores. In other words, for a stretch Entropik was losing more than its entire annual revenue, the classic cost of buying growth before the product could sell itself. Then came the pandemic, which forced consumer research online almost overnight, and forced Entropik to decide whether it was a lab company or a cloud company.

The turning point

The pivot from lab to self-serve SaaS is the single event that separates the two halves of Entropik’s story, and the numbers on either side make the case. Before it, in FY22, the company reported revenue of about ₹16.39 crore (₹15.3 crore from operations, per Entrackr) and a net loss of ₹17.7 crore. After it, in FY23, revenue rose to ₹47.53 crore, an increase of roughly 189% year on year, as compiled by thekredible from the company’s filings and corroborated by Inc42 at ₹47.4 crore.

That near-tripling is the payoff to the claim in the opening line. It was powered by moving the emotion-AI capability into cloud platforms brands could run themselves, and by leaning into the post-2020 surge in remote, online research. The catch, also promised in the opening, is that growth was not free: total expenses climbed to ₹74.6 crore in FY23, per Inc42, and the net loss widened to ₹27.2 crore, up about 52% year on year. The turning point proved the market; it did not, on this record, prove profitability.

The money behind it

Entropik has raised roughly $34 million across six rounds since incorporation, per Inc42’s funding data (Tracxn puts the figure slightly higher, near $35.5 million). The shape of the cap table, with periods:

• Seed, November 2016 — $200,000 (Inc42 funding data).
• Seed/pre-Series A, July 2018 — $1.1 million from Bharat Innovation Fund and IDFC-Parampara Capital (Inc42; BW Disrupt).
• Undisclosed round, December 2018 involving Inflexor Ventures (Inc42).
• Series A, September 2020 — ₹60 crore ($8 million, as reported), led by Alpha Wave Incubation (managed by Falcon Edge Capital), with Bharat Innovation Fund and IDFC-Parampara participating; the company was valued at about ₹300 crore post-money, roughly a 10x jump from about ₹30 crore at seed (Entrackr; YourStory).
• Grant, January 2021 — $25,000 via Flipkart Leap (Inc42).
• Series B, February 2023 — $25 million, led by Bessemer Venture Partners and SIG Venture Capital, with Trifecta Capital, Alteria Capital and Bharat Innovation Fund participating (Entrackr; BusinessWire; YourStory). The valuation for this round was not disclosed.

What each backer changed is visible in the trajectory. Bharat Innovation Fund provided the early institutional conviction and re-upped repeatedly. Alpha Wave Incubation’s Series A funded the shift to a cloud platform and delivered the 10x valuation step. Bessemer and SIG’s Series B, arriving after the FY23 revenue jump, was the capital meant to scale a proven self-serve motion internationally. The last publicly disclosed valuation remains the September 2020 figure of about ₹300 crore; any 2023 or later mark is not in the public record used here.

How it makes money

Entropik is a business-to-business SaaS company, so the money mechanics are recurring software revenue, not one-off sales. Broadly, per the company’s model:

• Money in: subscriptions and platform fees for Decode and Qatalyst, sold to brand, insights and product teams, typically as annual contracts that scale with seats, studies or respondents.
• Costs out: the heavy lines are people (engineering and data-science talent to build and maintain the emotion-AI models) and go-to-market spend (sales and marketing to win enterprise logos). In FY23, total expenses of ₹74.6 crore against revenue of ₹47.53 crore show cost still exceeding income (Inc42; thekredible).
• Where the margin sits: gross margin on software is structurally high once the models are built; the loss is an operating-leverage problem, not a per-unit one. The path to profit is spreading fixed R&D and sales costs over a much larger recurring-revenue base.
• The part people get wrong: Entropik is often described as a hardware or neuroscience firm because of the brainwave heritage. In practice its scalable products run on ordinary webcams and browsers; the “neuro” signal is software inference, which is what makes SaaS-style margins possible at all.

The numbers

The public filings, as compiled by Entrackr, Inc42 and thekredible, tell a consistent three-year story of fast revenue growth alongside widening losses. Figures are ₹ crore.

Reading across the three years:

• Revenue: ₹7.98 crore (FY21) to ₹16.39 crore (FY22) to ₹47.53 crore (FY23), a near-6x rise over two years (Entrackr; thekredible).
• Losses: ₹2.91 crore (FY21) to ₹17.7 crore (FY22) to ₹27.2 crore (FY23) — losses grew even as revenue accelerated (Entrackr; Inc42).
• Expenses: ₹74.6 crore in FY23, up about 118% year on year, outpacing revenue (Inc42).
• Availability: the most recent detailed public filing used here is FY23 (year ended 31 March 2023); verified FY24 and FY25 figures were not available in the public sources consulted for this piece, so they are deliberately omitted rather than estimated.

Where the money comes from

Entropik’s revenue mix is defined by product and by geography, though the company does not publish an audited segment table in the sources available. What the record supports:

• By product: two platform lines carry the revenue — Decode for consumer and market research, and Qatalyst for UX/user research — layered on the older Affect Lab emotion-AI foundation and, from 2024, a generative-AI layer marketed as Decode Copilot (company materials).
• By customer: large consumer brands and their agencies, with disclosed users spanning FMCG (Mondelez, Reckitt, P&G, Kellogg’s), food service (KFC), pharma (Cipla), insurance (Tata AIG) and media (Disney), per the company’s client list.
• By geography: Entropik operates as a global SaaS seller from a Bengaluru base, targeting international insights budgets rather than only the Indian market; the Series B was explicitly framed by investors as fuel for global expansion (BusinessWire).

The surprise in the mix is how little of the value now depends on the brainwave hardware that gave the company its “neuro” reputation. The scalable revenue rides on camera- and browser-based inference, which is exactly what lets a research idea born in a lab be sold as software worldwide.

The risks

Three concrete risks sit on this business, each with a clear mechanism:

• Persistent cash burn. With FY23 expenses of ₹74.6 crore against ₹47.53 crore of revenue and a ₹27.2 crore net loss (Inc42; thekredible), Entropik depends on external capital to fund growth. If losses do not narrow as revenue scales, the company needs another round on terms it cannot control, and the last disclosed valuation dates to September 2020.
• Regulation and trust in emotion AI. Inferring emotion from faces, eyes and voice is squarely in the path of tightening biometric- and AI-privacy rules globally. Restrictions on facial analysis or consent requirements could raise compliance costs or shrink addressable use cases; this is a mechanism risk, not a hypothetical, for any company reading biometric signals.
• Commoditisation by large AI vendors. As general-purpose AI models add vision and speech-sentiment features, Entropik’s differentiation must stay ahead of capabilities that hyperscalers can bundle cheaply. The defence is depth in research workflows and validated emotion models; the risk is that “good enough” emotion inference becomes a free feature elsewhere.

The takeaway

The transferable lesson from Entropik is about the distance between a true insight and a sellable product. The founding observation, that people buy emotionally and explain rationally, was sound from day one and never really changed. What changed, painfully and expensively, was the delivery: from a brainwave lab to a camera-and-browser SaaS platform that a brand team can run without a neuroscientist in the room. The revenue only tripled once the packaging matched the buyer. Founders often assume the hard part is the idea; Entropik’s filings suggest the hard, capital-hungry part is turning a correct idea into something that scales, and surviving the years of losses in between.

Frequently asked questions

What does Entropik Technologies do?

Entropik is a Bengaluru-based SaaS company that sells “Emotion AI” research software. Its platforms, Decode and Qatalyst, use facial coding, eye tracking and voice analysis (and brainwave data in its Affect Lab product) to help brands measure how consumers emotionally respond to ads, packaging and digital experiences.

Who founded Entropik and when?

Entropik Technologies was incorporated on 1 February 2016 in Bengaluru. It was co-founded by Ranjan Kumar, the CEO, and Lava Kumar, the Chief Product Officer, with the idea tracing back to a student project at IIT Kharagpur, per company profiles and YourStory.

How much money has Entropik raised?

Entropik has raised roughly $34 million across six rounds since 2016, per Inc42 (Tracxn cites about $35.5 million). Its largest round was a $25 million Series B in February 2023, led by Bessemer Venture Partners and SIG Venture Capital.

Is Entropik profitable?

No, not on the latest available public filings. For FY23 (year ended 31 March 2023) Entropik reported revenue of ₹47.53 crore and a net loss of ₹27.2 crore, with total expenses of ₹74.6 crore, per Inc42 and thekredible. Losses widened even as revenue nearly tripled.

What is Entropik’s valuation?

The last publicly disclosed valuation is about ₹300 crore post-money, set at the September 2020 Series A round (Entrackr, YourStory). The February 2023 Series B did not disclose a valuation, so no more recent figure is in the public record.

Sources

Figures are as of September 2026. Currency converted at $1 ≈ ₹96.0 as of 18 September 2026 (Trading Economics).

• Entrackr — “Entropik raises $25 Mn Series B round led by Bessemer and SIG” (February 2023)
• Entrackr — “Entropik raises Rs 60 Cr Series A round, valuation soars by 10X to Rs 300 Cr” (September 2020)
• BusinessWire — “Entropik… raises $25 million in Series B funding round led by Bessemer Venture Partners and SIG Venture Capital” (February 2023)
• YourStory — “AI analytics startup Entropik Tech raises $8 million Series A” (September 2020) and Series B coverage (February 2023)
• Inc42 — Entropik company, financials and funding profiles (accessed September 2026)
• thekredible — “Entropik records 2X rise in topline in FY23, losses widen by 52%” and Entropik financials (accessed September 2026)
• Tofler / MCA registry — Entropik Technologies Private Limited incorporation and company details (accessed September 2026)
• Tracxn — Entropik company and legal-entity profiles, headcount (accessed September 2026)
• BW Disrupt — “Entropik Tech Raises $1.1M from Bharat Innovation Fund & IDFC-Parampara Capital” (2018)
• Trading Economics — USD/INR reference rate, 18 September 2026

Found an error? Write to us and we’ll correct it in the open, dated, on the piece.