Startup Deep Dive : Augnito — the profitable medical voice AI built on a 20-year NHS transcription business

Augnito India Private Limited posted a net profit margin of 36.7% in the year to March 2025, a number most venture-funded AI companies never reach, while its revenue grew 375% in FY24 and another 54.3% in FY25 (Tofler, MCA-derived). That profit did not come from a VC war chest. The medical voice engine now running in more than 500 hospitals across 25-plus countries was paid for, year after year, by the cash of a 20-year-old NHS transcription business its founders started as teenagers.

This is the contradiction worth sitting with. Augnito is an AI startup that behaves like an old cash business: patient, self-funded for most of its life, and profitable before it took a real institutional cheque. It sells software that lets a doctor speak a clinical note and have it typed, structured and filed in seconds. But the story underneath is about a transcription BPO called Scribetech that spent almost a decade turning its own back office into a product, and then sold that product back to the hospitals it once served as a vendor. The rest of this piece traces how a Mumbai-registered company with roots in London’s health service reached a 99% out-of-the-box accuracy claim, a 375% revenue jump, and a rare early profit.

Quick facts

What they do

Augnito sells voice AI built for one job: getting a clinician’s spoken words into the medical record accurately, without training the software on each user’s voice first. A radiologist can dictate a report, a physician can speak a consultation note, and the text lands inside the hospital’s existing electronic record system. The company sells to hospitals, clinics and health-software vendors rather than to consumers, on enterprise subscription terms. Its three main products cover the front end (dictation), the back end (developer APIs), and the newer ambient layer that listens to a doctor-patient conversation and drafts the note automatically.

The origin

The founding insight predates Augnito by almost two decades. In 2004, Rustom Lawyer started Scribetech with Shiraz Austin, a clinical documentation company that transcribed medical notes for the UK’s National Health Service and independent hospitals. Lawyer was 19. Scribetech’s clients came to include NHS trusts such as Chelsea and Westminster, Kingston and hospitals in the Birmingham area. For years the business was labour: skilled transcriptionists turning dictation into structured reports across dozens of specialties.

The insight that became Augnito was simple and uncomfortable for a transcription company to act on. If software could do accurately what a room of transcriptionists did, the same firm that sold the human service could sell the machine that replaced it. Lawyer, who spent years on executive programmes at Harvard Business School and MIT, kept circling one gap: Indian healthcare barely used digital documentation at the point of care, and doctors everywhere were drowning in typing. Rather than defend the BPO, Scribetech began building speech recognition tuned to clinical language, funded quietly out of its own profits. Augnito was the result, launched in 2020 as a product covering more than 50 specialties.

The struggle years

The hard part was not the idea; it was the decade of building with no outside money and a business model that pointed at cannibalising the founders’ own revenue.

• Building the engine inside a services company (roughly 2011 onward): the speech technology started as an internal efficiency tool for Scribetech’s transcription work, not a product. Turning a cost-saving hack into medical-grade software that hospitals would trust took years, not quarters.
• Bootstrapping through the build: the founders took no institutional venture capital during this period and reinvested Scribetech’s profits into R&D. That kept them alive but slow — a self-imposed constraint that meant every rupee of engineering had to be earned first.
• Selling against the incumbent and against habit: global clinical speech recognition was dominated by long-established Western vendors, and the harder problem was persuading doctors — especially in India, where point-of-care digital documentation was thinly adopted — to change how they work. As late as November 2023 the CEO was still describing the company as only then “closing in on our first funding round,” a sign of how long it ran on its own cash.

The turning point

The turn came when the product stopped being a Scribetech side-project and started being adopted at scale by hospital groups that had every reason to be sceptical. The clearest single marker is Apollo Hospitals, Asia’s largest hospital group, which moved from customer to investor.

The numbers on either side of that shift tell the story. In its published Apollo Hospitals case study, Augnito reports a 46% productivity increase for clinicians using the tool, with physicians saving roughly 44 hours a month on documentation. On deployment, the company went from “more than 375 hospitals across more than 25 countries” (June 2024) to “more than 500 hospitals” and about 15,000 clinicians in interviews through 2024 and 2025. On the financials, the Indian entity’s revenue rose 375% in the year to March 2024, then a further 54.3% in the year to March 2025 — while staying profitable. A product built to make one transcription firm faster had become infrastructure for hundreds of hospitals.

The money behind it

Augnito’s cap table is unusual for an AI company: small, late, and anchored by a customer. The funding shape:

• Total raised: about $7.2 million (roughly ₹69 crore) across four rounds, from 15 investors (Tracxn, as of 2026). That is modest for a company at this deployment scale, reflecting the years of self-funding.
• Apollo Hospitals — anchor and customer. Apollo, the group whose own doctors use the product, made its first investment in Augnito’s seed round dated November 2022 (Tracxn). A hospital chain backing the vendor it deploys is a strong signal of real usage.
• Alpha Care — the second institutional investor alongside 13 angels (Tracxn).
• Ishaat Hussain — credibility hire. The former finance director of Tata Sons sits on Augnito’s advisory board, a governance signal rare at this size.
• Dharmendra Ghai — strategic, Gulf-focused. On 11 June 2024 Augnito announced a strategic investment from Ghai, a Qatar-based digital-health investor, to accelerate deployment in the Middle East (amount undisclosed).
• Latest round dated 2 December 2025 (Tracxn). Per-round amounts and the post-money valuation are redacted or masked on the tracker databases, so this piece does not state them.

The honest caveat: round-by-round rupee figures on startup trackers for Augnito appear obfuscated and do not reconcile with the reported $7.2 million total, so only the total, the named backers and the round dates are treated as reliable here.

How it makes money

Augnito is a business-to-business software subscription. The money in, and where the margin sits:

• Per-seat SaaS subscriptions: hospitals and clinics pay recurring fees, typically scaled by the number of clinician seats, EMR integrations and modules chosen. Pricing is quote-based rather than public-tiered (SaaSworthy, G2).
• Three product lines, one engine: Augnito Spectra (front-end dictation for individual clinicians and departments), Augnito Voice Services (an SDK/API sold to health-software vendors who embed the speech engine), and Augnito Omni (the ambient AI scribe using generative AI). Selling the same core engine three ways spreads the R&D cost across more revenue.
• Where the margin sits: once the speech models are built, additional seats carry low marginal cost — the classic software economics that show up in Augnito India’s FY25 operating margin of 43.6% and net margin of 36.7% (Tofler). The heavy spend is R&D and enterprise sales, not per-customer delivery.
• The part people get wrong: this is not a pure “AI startup burning to grow” story. The heritage in Scribetech’s services business meant Augnito priced for profitability from early on rather than buying market share at a loss.

The numbers

The Indian operating entity, Augnito India Private Limited, files with the MCA. Figures below are from Tofler and thecompanycheck (MCA-derived); the FY23 and FY25 rupee values are approximations derived from the reported year-on-year growth rates and are labelled as such.

Two things stand out. First, the growth rate is decelerating in percentage terms (375% then 54.3%) but off a larger base, which is normal as revenue scales. Second, and more unusual, the company is profitable at a healthy margin while still growing fast — a combination Indian tech startups rarely show this early. Paid-up capital stood at about ₹43.1 crore (Tofler), reflecting capital infused into the Indian entity.

Where the money comes from

Augnito is a genuinely cross-border business, and the surprise is how much of its identity and demand sits outside India even though the operating company is Mumbai-registered.

• India: Apollo Hospitals, Max Healthcare, Fortis Healthcare and NM Medical are named users (pharmabiz, Nov 2023).
• United Kingdom: the NHS lineage runs directly from Scribetech’s two decades of transcription work for NHS trusts; the UK remains a core market.
• Middle East / GCC: the fastest-growing front, with partners including King’s College Hospital London (Dubai), Prime Healthcare, Dallah Hospitals, King Abdulaziz University Hospital, RAK Hospital and ATTIEH Medico. The 2024 Dharmendra Ghai investment was explicitly aimed at this region.
• United States: customers such as Physician One (pharmabiz, Nov 2023).

The surprise: a company that files its accounts in Mumbai draws demand from more than 25 countries, and its deepest institutional roots — the NHS relationships that seeded the whole thing — are British. Augnito is “Made in India” voice AI whose founding customer base was the UK health service.

The risks

• Big-tech and incumbent competition: clinical speech recognition is contested by long-established Western vendors and, increasingly, by general-purpose large language models and ambient-scribe startups. Augnito’s 99% out-of-the-box accuracy claim (company-stated) is a moat only as long as rivals cannot match it on clinical vocabulary and accents — and the barrier to entry is falling as foundation models improve.
• Concentration and reference-customer dependence: Apollo Hospitals is simultaneously a marquee customer, a case study and an investor. That alignment is powerful for credibility, but it means a large share of the India narrative rests on a small number of flagship accounts; losing one would hurt disproportionately.
• Regulatory and clinical-safety exposure: the product touches the medical record. Errors in a dictated note are not cosmetic — they can affect care. As health-data rules tighten across India, the UK and the GCC, compliance cost rises, and any high-profile accuracy failure in a live clinical setting is a reputational risk the company cannot fully control.

The takeaway

The transferable lesson from Augnito is that a services business can be the best possible R&D lab for a product — if the founders are willing to build the thing that eats their own revenue. Scribetech had two decades of clinical dictation data, real NHS customers, and a painful understanding of exactly where transcription was slow and expensive. Instead of protecting that BPO, Lawyer and Austin used its profits and its domain knowledge to build the software that made it obsolete, and priced the software to make money from day one rather than chasing scale at a loss. Most incumbents defend the old business until a startup with none of their data does it to them. Augnito’s edge was choosing to be that startup to itself.

Frequently asked questions

What does Augnito do?

Augnito makes medical voice AI that turns a clinician’s speech into accurate, structured text inside a hospital’s electronic record system. Its products are Augnito Spectra (dictation), Augnito Voice Services (a developer API) and Augnito Omni (an ambient AI scribe). It sells to hospitals, clinics and health-software vendors, not consumers.

Who founded Augnito and what is its link to Scribetech?

Augnito was co-founded by Rustom Lawyer (CEO) and Shiraz Austin, who in 2004 had already founded Scribetech, a UK clinical transcription company serving the NHS. Augnito’s speech engine grew out of Scribetech’s own transcription work and was funded for years by that business’s profits before launching as a product in 2020.

Is Augnito profitable, and how much revenue does it make?

Augnito India Private Limited reported operating revenue of about ₹10.7 crore in FY24 (up 375% year on year) and roughly ₹16.5 crore in FY25 (up 54.3%), with a net profit margin of about 36.7% in FY25, according to MCA-derived data on Tofler and thecompanycheck. That makes it profitable, which is unusual for an AI startup at this stage.

How much funding has Augnito raised and who are the investors?

Augnito has raised about $7.2 million (roughly ₹69 crore) across four rounds from 15 investors (Tracxn, 2026). Its institutional backers are Apollo Hospitals and Alpha Care, and its advisory board includes Ishaat Hussain, former finance director of Tata Sons. In June 2024 it also announced a strategic investment from Qatar-based investor Dharmendra Ghai.

Where is Augnito used?

Augnito says its software is used in more than 500 hospitals across over 25 countries by roughly 15,000 clinicians. Named users include Apollo Hospitals, Max Healthcare and Fortis in India; NHS trusts in the UK; and hospitals across the Middle East such as Prime Healthcare, Dallah Hospitals and RAK Hospital.

Sources

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

• Tofler — Augnito India Private Limited company and financial profile (MCA filings; incorporation, directors, paid-up capital, FY25 growth and margins), September 2026
• thecompanycheck — Augnito India Private Limited, FY24 revenue and YoY growth, September 2026
• Tracxn — Augnito company profile, funding rounds and investors (total raised, round dates, institutional and angel investors), 2026
• Crunchbase — Augnito organization and funding profile, 2026
• Pulse2 — interview with co-founder and CEO Rustom Lawyer (products, accuracy, specialties, hospitals, clinicians, Apollo case metrics), 2024
• Pharmabiz — “Augnito aims to capture 30% of the healthcare voice AI market… 300% year on year growth” (deployment, customers, market size), November 2023
• PR Newswire — “Augnito Announces Strategic Investment to Accelerate Digital Health Innovation” (Dharmendra Ghai, 375+ hospitals, GCC partners), June 2024
• IndiaAI (indiaai.gov.in) — Augnito profile and Scribetech-to-AI background, 2022–2024
• YourStory — coverage of Augnito’s AI documentation tool and 2020 launch, September 2020
• SaaSworthy and G2 — Augnito Spectra pricing model (quote-based enterprise SaaS), 2026

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