Factify’s $73M Bet: Turning Static PDFs Into Intelligent, API‑Like Documents

For most enterprises, the humble PDF or .docx file is still the backbone of daily work: contracts, investment memos, policies, filings, HR paperwork. Yet for AI systems and compliance teams, these same files are frequently black boxes—opaque, hard to govern, and easy to lose track of once they leave core systems.

Tel Aviv–based startup Factify is emerging from stealth with a $73 million seed round to attack that problem at its root: the document format itself. Rather than adding another layer of workflow or collaboration on top of PDFs and Word files, Factify wants to replace them with a new, intelligent document standard that behaves much more like an API than a file.

For enterprise technology leaders, the company’s core claim is simple but consequential: static file formats are now a liability in an AI-driven world, and the organizations that move to “intelligent documents” first will be better positioned to manage risk and unlock AI value from their records.

Why Today’s Document Formats Are Hitting a Wall

Factify’s founder and CEO, Matan Gavish—a tenured computer science professor and Stanford PhD—frames the current situation as a structural limit, not just an inconvenience. The core formats the industry relies on were never designed for modern expectations around security, auditability, or AI-readiness.

Despite decades of software progress, the basic formats that matter most in business have barely changed. The PDF itself, Gavish notes, dates back to his elementary school years. Yet it remains the de facto standard for anything “final” in business and government: contracts, tax filings, formal disclosures, compliance documents, and more.

The problem, as Factify sees it, is that these artifacts are still treated as static blobs:

• Once exported, documents largely lose track of who sees them, where they go, and what version is authoritative.
• Security controls are usually perimeter-based—gateways, VPNs, email rules—rather than embedded within the document itself.
• For AI systems, PDFs and many legacy files are essentially images or loosely structured containers, demanding error-prone extraction before they can be used.

This is the context in which investors, led by Valley Capital Partners and including AI leaders such as former Google AI chief John Giannandrea, are backing Factify’s attempt to redesign the digital document itself.

How We Got Here: From Word Files to PDFs to Cloud Docs

The scale of the challenge is rooted in history. Factify points out that the evolution of digital documents has not been a simple story of linear replacement. Instead, different document “species” emerged for different jobs: creation, distribution, and collaboration. That speciation is now what fragments enterprise workflows.

Three key eras define today’s status quo.

The era of files: Microsoft Word and binary .doc

In the 1980s and 1990s, documents were tightly coupled to specific software and hardware. A file created in WordPerfect on DOS could be meaningless on a Macintosh. Microsoft Word, drawing from WYSIWYG innovations at Xerox PARC, broke through by riding the dominance of Windows. The binary .doc format quickly became the default for editable, professional documents.

But .doc files were complex artifacts, effectively “memory dumps” tuned for constrained hardware. This design led to familiar operational headaches for IT and legal teams: file corruption, version incompatibilities, and even privacy leaks where “deleted” text remained buried in the binary data.

The era of digital ‘stone’: PDF as final record

The next major shift came with Adobe’s PDF. Originating in the early 1990s from the Camelot Project, PDF was not built for writing but for viewing. It was intended as a “digital envelope” that would look identical on any screen or printer.

By leveraging the PostScript imaging model, PDF locked in layout fidelity. Text and graphics were placed at precise coordinates—excellent for presentation, less so for downstream machine processing. Adoption accelerated when Adobe made Acrobat Reader free in 1994. The format became global “digital concrete,” used wherever visual exactness and finality mattered: contracts, government forms, long-term archives.

Today, there are an estimated three trillion PDFs in circulation. As Gavish jokes, some people see PDFs more than they see their own kids. For enterprises, this represents both a massive installed base and a significant technical and governance burden.

The collaborative cloud docs era

In 2006, Google Docs pushed documents into the browser and into real-time collaboration. Using Operational Transformation algorithms, Google enabled multiple people to edit the same text stream simultaneously, changing how teams worked together.

The paradigm shifted from “sending files” to “sharing links.” Google Workspace, now with billions of users across consumers and education, helped normalize the idea of the document as a living process rather than a static object. Yet even here, when a document needs to be shared externally, signed, or archived, it usually ends up back in PDF form.

The net result is a three-part ecosystem most enterprises will recognize: cloud docs for drafting, Word for formatting, and PDFs for signing and record-keeping. Factify argues this fragmentation is now unsustainable.

Why Fragmentation Is Now a Business Risk

In day-to-day operations, this multi-format ecosystem introduces constant friction. For CIOs, CTOs, and operations leaders, it also represents a growing risk surface.

Today, a typical document lifecycle might look like this:

• A policy or memo is drafted collaboratively in a cloud document editor.
• It is then exported to Word for formatting or specific styling requirements.
• Finally, it is rendered as a PDF for distribution, e-signature, or filing.

Each step potentially splits versions, permissions, and context. Once a PDF leaves core systems by email or download, visibility and control often drop sharply. Enterprises rely on a patchwork of systems—DLP, CASB, email gateways, IRM—to compensate, but the document itself remains unaware of its own history, owners, or intended boundaries.

Factify summarizes this bluntly: “The problem is not the document. It is everything around it.” When a PDF leaves your system, versions drift, access becomes opaque, and the organization loses a clear line of sight into what is happening with what may be a mission-critical artifact.

In isolation, this is an operational annoyance. At scale—across thousands or millions of contracts, HR records, financial documents, and regulatory filings—it becomes a systemic governance and risk issue.

From Files to ‘Documents With a Brain’

Factify’s core wager is that, in the era of AI, treating documents as passive files is no longer viable. AI systems need structured, verifiable data. Static PDFs and legacy formats force every downstream system to guess.

When an AI “reads” a PDF, it often relies on text extraction or optical character recognition (OCR), working from what is effectively a digital picture. That process is lossy and error-prone, and it rarely captures the intent, history, or permission model attached to the original document inside the enterprise.

Gavish describes Factify’s approach as building “intelligent infrastructure” for documents. Instead of acting like dead storage objects, documents in Factify’s proposed standard carry an embedded “brain” with three main attributes:

• Unique identity: Each document is a first-class, addressable entity—not just another file in a folder.
• Live permission system: Access, sharing, and forwarding rules live with the document itself, not just in the surrounding systems.
• Immutable audit log: A complete, portable history of who accessed the document, how, and when, travels with it.

Under the hood, Factify says it has written a new document format that “supplants the PostScript” approach underlying PDFs. On top of that, it has built a new data layer that treats documents as first-class citizens, always available inside the organization and potentially extendable outside it.

Conceptually, this reframes the question from “Where is the file?” to “What can this document tell me about itself?” A Factify document, in the company’s own framing, should be able to answer queries like:

• Who has seen you?
• When do you expire?
• Are you the most up-to-date version?

For enterprise architects, this is less about a single product feature and more about a shift in abstraction. Documents become API-like objects with state, permissions, and history, rather than inert blobs that must be managed by external systems.

Files as Liabilities, APIs as Assets

Gavish draws a sharp line between traditional files and what Factify is trying to build. In his view, files are increasingly “brick-like”: once created, they accumulate in storage, must be guarded, and give little back in operational insight or intelligence.

Factify frames this as a liability: files can be lost, exfiltrated, or misused, and they rarely surface their own risk. They also resist the kind of granular, programmatic interrogation that modern security and AI systems need.

By contrast, APIs in modern software architectures are treated as assets. They expose clear interfaces, can be monitored, versioned, and governed, and are composable across products and services.

Factify’s pitch is to bring that API-like model to documents themselves. Instead of being static end points in workflows, documents become active objects that can be queried and integrated more cleanly into enterprise systems—security platforms, analytics, AI assistants, and line-of-business applications.

Practically, this could change how enterprises think about compliance checks, records management, and AI integration over time. Rather than performing repeated one-off extractions from PDFs or .docx files, systems could plug directly into an authoritative document layer that exposes state and metadata in a more structured, durable way.

Driving Adoption Without Forcing Users to Change

However, history is full of attempts to dethrone the PDF that never took hold, such as Microsoft’s XPS format. Most failed because they required significant user behavior change or new tooling for limited perceived gain.

Gavish is explicit about this risk. He argues that enterprise software succeeds only when it respects two hard truths: “people don’t care, and no one changes.” In practice, that means the new standard cannot demand that everyday users learn a new interface, workflow, or visual paradigm.

To address this, Factify emphasizes deep backward compatibility and familiarity:

• A “Factified” document can be rendered to look exactly like a PDF, with conventional page breaks, margins, and layouts.
• End users can continue to interact with documents in familiar ways, while the underlying format quietly carries identity, permissions, and audit logic.

Gavish envisions adoption starting with very specific, high-stakes use cases—such as an executive who wants to ensure that investment memoranda or employment agreements cannot be forwarded beyond a defined circle of recipients. In this model, a leader could simply instruct a team that certain classes of documents “are going to be Factified” while everything else stays as-is.

The expectation is that teams will experience immediate value around control, auditability, and clarity of versioning for those critical documents. Over time, that may encourage wider adoption without a heavy-handed migration mandate.

This approach acknowledges that while formats can change under the hood, user habits are slow to move. For enterprise adoption, minimizing visible disruption is often as important as the underlying technical advance.

What’s Next for Factify — and What It Means for Enterprises

Factify plans to use its $73 million seed funding to continue what Gavish describes as a “heavy engineering lift.” The company is rebuilding three layers from the ground up:

• The document format itself, as an alternative to PDF’s PostScript roots.
• The data layer that treats documents as persistent, queryable entities.
• The application layer through which organizations will create, manage, and interact with these new documents.

In parallel, Factify is setting up a major operational hub in Pittsburgh to support its expansion in the United States, signaling a focus on large enterprise and institutional customers.

Strategically, the company is not positioning itself as a competitor to collaborative suites like Google Docs. Instead, its ambition is to define the “immutable record of the future”—a generic, domain-agnostic standard for digital truth. Gavish points to the PDF’s dominance in tax filing as an example of what victory looks like: institutions and governments broadly agreeing that a given format is the one acceptable way to represent a final, authoritative record.

Factify similarly aims for a standard that is not tied to a single vertical such as healthcare or insurance, but that could operate as a universal substrate for documents of any kind.

For enterprise technology leaders, several implications stand out:

• AI-readiness of records: If documents become inherently structured, addressable objects with stable identities and histories, AI initiatives across legal, finance, HR, and compliance can rely less on ad hoc parsing and more on a robust underlying substrate.
• Risk and governance: Document-centric audit trails and embedded permissions could reduce reliance on purely perimeter-based controls and offer clearer evidence trails for regulators and auditors.
• Migration strategy: With trillions of existing PDFs and other static files, any path forward will inevitably be gradual. Factify’s compatibility-focused approach suggests a coexistence phase where a subset of critical documents migrates first.

Whether Factify can turn its “megalomaniac vision,” as Gavish calls it, into an industry standard remains to be seen. But the company has put a large stake in the ground: in a world where AI and compliance expectations are accelerating, the static file may no longer be good enough as the authoritative record of business.