For decades, document redaction has meant one thing: a human reviewer, a stack of files, and a black marker — or its digital equivalent. It was slow, expensive, and prone to catastrophic error. The kind of error that made headlines. Now, PTFS is building AI technology that is making that model look like a relic.
Progressive Technology Federal Systems (PTFS), founded in 1995 and headquartered just outside Washington, D.C., has been quietly building content management software for government agencies since before most people knew what a content management system was. Their flagship platform, Knowvation — originally called ArchivalWare when it launched in 2004 — has grown into a full enterprise content services suite. But it’s Knowvation DX, the platform’s dedicated redaction and declassification module, that’s causing the most uproar.
The Problem with Doing It by Hand
To understand why Knowvation DX matters, you have to appreciate just how broken the old process was. Traditional redaction required skilled reviewers to manually read through large volumes of documentation like scribes of ancient times, searching for sensitive terms, classified concepts, or personally identifiable information (PII) that couldn’t be released publicly. It was a labor-intensive and dangerous job.
PTFS’s own documentation puts it plainly: the process was “monotonous and tiring, conducive to human fatigue and an increased likelihood of error.” Reviewers weren’t incompetent; they were just human. And when you’re reading your five hundredth document of the week, looking for buried Social Security numbers or classification markers, things slip through.
The legal stakes are severe. Under the Freedom of Information Act (FOIA), enacted in 1966, federal agencies are required to release records upon request unless they fall under specific exemptions, including defense information, personal privacy, law enforcement techniques, and others. Release too much, and you risk national security or individual privacy. Release too little, and you’re not meeting your legal obligations. Getting it wrong, in either direction, can have serious repercussions.
What Knowvation DX Actually Does
Knowvation DX approaches the problem through semi-automation — not total autopilot, but a machine-assisted workflow that improves on what a human reviewer could accomplish. The system combines natural language processing, fuzzy text search, and semantic indexing to flag sensitive content across documents, including content that might bypass a simple keyword search due to varied phrasing or OCR errors.
Most importantly, the platform doesn’t just black out text. It cleanses information across all document layers: the visual content, the OCR text layer, and the metadata. This is important because some of the most damaging redaction failures in history have come from PDFs where the “hidden” text was simply covered by an opaque box — leaving the underlying data fully extractable by anyone who knew how to look. Knowvation DX’s PDF Sanitizer addresses this by auditing the internal object structure of PDF files and removing sensitive content from embedded layers, attached files, and comments that are invisible to the average reader but accessible to a knowledgeable reader.
According to PTFS, the platform delivers reviews 10 times faster than a manual process — a figure that, in government procurement terms, translates directly into cost reductions and faster FOIA response times.
Deployed Where It Matters
Knowvation DX isn’t just a proof of concept. It’s been used by mission critical government organizations, including Department of Defense components and intelligence agencies. Knowvation CSP supports over 45 Government agencies. PTFS also operates an internal digitization service bureau that has provided classified and unclassified conversion services to more than 100 government organizations — meaning the company isn’t just selling software; it’s embedded in the actual workflow.
The platform meets a serious roster of compliance standards: FedRAMP, NIST, DoD Impact Level requirements, HIPAA, and NARA UERM. It can run in PTFS-hosted environments, on-premises, or in Government cloud (Govcloud) operated infrastructure — flexibility that matters as agencies increasingly migrate to unclassified, Secret and Top-Secret cloud environments.
PTFS has also worked on the Air Force’s declassification of historical microfilm. This project involved preserving deteriorating sensitive media and running it through Knowvation DX for redaction before public release. That kind of mission — salvaging irreplaceable records while ensuring they can be released safely — is exactly where the technology earns its keep.
Why It’s Different
PTFS describes Knowvation DX as “the only product on the market focused on declassification automation and precise redaction using a semi-automated approach.” That’s a bold claim, but the specificity is telling. The platform’s flexible workflow engine can be reconfigured in response to changes in FOIA rules, privacy act requirements, or declassification guidance without rebuilding the system from scratch. And its web services architecture means that new algorithms and capabilities can be plugged in over time.
The “human in the loop” framing is deliberate. Knowvation DX isn’t trying to remove reviewers from the process; it’s trying to ensure they spend their time on genuinely difficult judgment calls rather than manually hunting for obvious hits across thousands of pages of documents.
That’s a meaningful distinction — and probably the reason the platform has found as much traction as it has in an industry where the cost of getting it wrong can’t be measured in dollars alone.
Sources: PTFS.com (Knowvation DX product page, white papers, solutions overview, March 2026 blog post); AWS Marketplace listing for Knowvation CSP; Global Business Leaders Magazine (April 2022); PTFS press releases via GIS User and The Robot Report.
Leave a Reply