Apryse
The company formerly known as PDFTron sells the invisible machinery behind millions of documents - SDKs that let any app create, view, edit, sign, and mine PDFs without ever touching Adobe.
Somewhere today, you probably opened a PDF inside an app that was not Adobe Acrobat. A loan document in your banking portal. A boarding pass in an airline app. A contract that appeared in your browser, ready to sign, without a single download. It felt like nothing. That nothing is a business - and one of the more interesting ones in enterprise software belongs to a company most people have never heard of: Apryse.
Apryse does not sell document software to you. It sells document software to the people who build your software. Its products are SDKs - software development kits - that let engineering teams drop a full document engine into their own applications: viewing, editing, annotating, redacting, converting, e-signing, scanning, and extracting data from PDFs and more than 100 other file formats. The customer list runs to over 20,000 organizations, including 85 of the Fortune 100. The developer base, swollen by acquisitions, passed one million users. And in 2025, its private equity owner Thoma Bravo was reported to be exploring a sale at a valuation above $3 billion.
Not bad for a company whose core product category is, by any honest description, plumbing.
Twenty-five years of being PDFTron
The company was founded in Vancouver in 1998 by Catherine Williams and Ivan Nincic, five years after Adobe invented the PDF itself. The name back then was PDFTron, and the pitch was simple and technical: a cross-platform SDK that rendered complex documents with speed and pixel-level accuracy. No consumer brand, no app store presence, no marketing sizzle. Just an engine that worked, sold to developers who had tried to build their own and learned why they shouldn't.
That is the underrated part of this story. A PDF viewer sounds like a weekend project until you actually attempt one. The format has accumulated three decades of edge cases: embedded fonts, layered annotations, digital signatures, form fields, redaction that has to actually destroy the underlying text rather than just paint over it, accessibility tagging, and files produced by thousands of different generators, many of them broken in creative ways. Teams that start building a document engine tend to discover the true scope somewhere around month six. PDFTron's business was being the "buy" that beats "build."
For two decades the company grew quietly and profitably as a founder-led boutique - first the developer standard on Windows and Linux, then expanding into mobile and, crucially, the browser. Its WebViewer product put the entire engine into JavaScript, rendering and editing documents client-side. That architectural choice would later become its sharpest commercial weapon.
The roll-up with a thesis
The modern chapter starts with money. In May 2019, Boston growth firm Silversmith Capital Partners put $71 million (CAD $95 million) into PDFTron - the company's first outside capital in two decades. In 2021, Thoma Bravo, the software-focused private equity giant, took a majority stake. What followed was one of the more deliberate acquisition runs in developer tooling: thirteen deals, each one a step in a document's life.
Xodo brought a free consumer PDF editor used by millions - and with it, a mass-market proving ground for the engine. Solid Documents brought best-in-class PDF-to-Office conversion. eversign became Xodo Sign, the e-signature play. iText, one of the most widely used open-source PDF libraries in Java history, brought an enormous installed base of developers. LEAD Technologies, acquired in March 2024, added AI-powered document and imaging toolkits. Then in July 2025 came a double: Scanbot SDK, a German mobile-first specialist in barcode scanning and OCR, and Accusoft, a veteran imaging and forms-processing firm out of Florida.
In 2023, the name changed. PDFTron - a brand with a quarter century of developer trust attached - became Apryse. Rebrands invite mockery, but this one had logic: the portfolio had outgrown both the PDF and the single product the old name described. The company that emerged sells itself as the SDK for the entire document lifecycle, from a warehouse worker's phone camera to an archive server, with AI-driven extraction increasingly in the middle.
Who actually buys this
Two kinds of customers, broadly. The first is software companies - vertical SaaS vendors in construction, legal, insurance, real estate, and healthcare whose users live in documents all day. When a construction platform lets a foreman mark up a blueprint, or a legal tool lets an associate redact a filing, there is a decent chance an Apryse engine is underneath, wearing the vendor's branding. The SDK is white-label by nature; Apryse's best work ships under other people's names.
The second is the enterprise itself: banks, insurers, hospital systems, airlines, government agencies. These buyers care about a specific architectural detail that doubles as Apryse's cleanest differentiation - client-side processing. WebViewer renders and edits documents in the user's browser; the mobile SDKs work on-device. Files never have to be shipped to a third-party cloud to be viewed or signed. For an industry where a leaked document is a regulatory event, "your data never leaves the building" is not a feature bullet. It is the reason the deal closes. The ISO 27001 and SOC 2 certifications do the rest of the paperwork.
The anti-Adobe playbook
The obvious question about any PDF company is the Adobe question. Apryse's answer is to not fight that fight at all. Adobe sells finished applications and cloud services to end users; Apryse sells raw components to developers, one layer down the stack. Acrobat is a destination. Apryse is an ingredient. A product team that wants document features inside its own app - without bouncing users out to another program, and without Adobe's licensing - is Apryse's natural customer, not Adobe's.
The real competitive set is other component vendors: Foxit, Nutrient (the company formerly known as PSPDFKit), ComPDFKit, Aspose, Syncfusion, and open-source options like Mozilla's PDF.js for teams willing to accept lighter rendering fidelity. Against these, Apryse argues breadth and consolidation - one commercial relationship covering web, mobile, and server, plus generation, signing, capture, and extraction, instead of five vendors stitched together. Whether that breadth is a moat or just a bundle is the kind of question that gets debated in a sale process. The reported numbers suggest the market has an opinion: over $100 million in annual EBITDA, growing more than 20 percent a year, per 2025 press accounts of the exit exploration.
The business model, and the free app doing double duty
The revenue engine is commercial licensing: annual SDK subscriptions and enterprise agreements sold to companies that embed the technology, with pricing that scales alongside usage and deployment. It is a model with famously high switching costs - once a document engine is woven through your product, ripping it out is open-heart surgery - which is precisely the quality private equity prizes and the reason analysts covering the sale process talked about premium multiples.
Then there is Xodo, the oddball in the portfolio. It is a free consumer PDF editor with millions of users, most of whom will never know the parent company's name. Strategically it earns its keep twice: as a freemium product with its own subscription tier, and as a permanent public stress test of the engine. Every crash that doesn't happen in Xodo is a sales asset when a Fortune 100 evaluation team asks about reliability.
Where this goes
Apryse today is run from Denver by CEO Cassidy Smirnow, who arrived in 2022 from insurance-software firm Vertafore, with offices across North America and the UK and the founders' original engineering culture still visible in the product line. The strategic bet is that the document is not going anywhere - contracts, claims, invoices, and medical records remain the load-bearing paperwork of the economy - but that the work done on documents is changing from human reading to machine extraction. Hence LEAD's AI toolkits, hence Scanbot's edge capture, hence the steady drumbeat of "intelligent document processing" in the company's roadmap language.
It is a wager on a thirty-year-old file format staying central to global business, made by a company that has been quietly right about that wager since 1998. If Thoma Bravo's reported $3 billion price tag holds, it would rank among the larger outcomes ever for a developer-tools company - achieved without most of the world ever seeing the product. The machinery stays invisible. The machinery is the point.