Daniele MoltisantiPrincipal AI Engineer & AI Strategy Lead
Connect
RedactGuardDesktop · Korgis
Product family
RedactGuardOne privacy workflow. Two local execution boundaries.
PRODUCT · DESKTOP · KORGISExperimental privacy tool

RedactGuard Desktop

Detect configurable PII locally, review every finding, and export only the minimized document while keeping reusable local execution behind Korgis.

Desktop workflowKorgis backboneHuman review
See the workflow ↓
RedactGuard Desktop review screen showing detected sensitive data before redaction
Detect locally → review → redactDesktop · Korgis
The problem

Sensitive data is not a fixed list.

What must be protected changes by document, industry, organization and workflow. That makes detection both a privacy problem and a maintenance problem.

01

Manual review is slow

Long documents make anonymization repetitive, expensive and inconsistent.

02

Static rules are hard to adapt

New PII often means new regex, mappings, exceptions and application logic.

03

Cloud AI changes the privacy boundary

The original sensitive document must leave the machine before it can be analyzed.

The hidden cost: every change in what counts as sensitive can become a software change.
The desktop solution

Own the document workflow.
Delegate local execution to Korgis.

RedactGuard owns the product problem — import, PII policy, review, redaction and export — while Korgis provides the reusable local AI execution boundary underneath it.

The result is a clearer separation between application workflow and local inference infrastructure, without making cloud AI the default path for sensitive content.

1
Define PIIUse a profile or add custom sensitive categories.
2
Execute locallyKorgis owns the local model execution path.
3
ReviewThe model suggests. The user confirms.
4
ExportOnly the minimized result moves to the next workflow.
Product proof

Load. Review. Export.

The desktop workflow stays visible and reversible until the user confirms the final redactions.

RedactGuard — Choose the document and profile
1 · LOAD

Choose the document and profile

Start from a built-in domain profile or a custom detection definition.

RedactGuard — Inspect every finding
2 · REVIEW

Inspect every finding

Keep, remove or adjust what the local model identified as sensitive.

RedactGuard — Share the minimized result
3 · EXPORT

Share the minimized result

Export only after the redaction choices have been explicitly reviewed.

Under the hood

Same product problem. Desktop execution boundary.

RedactGuard keeps document-specific responsibilities in the application and uses Korgis as the local AI backbone, so product logic and inference infrastructure do not collapse into one code path.

Same product problem. Desktop execution boundary. — RedactGuard architecture

RedactGuard Desktop

Owns import, document processing, PII policy, review, redaction and export.

Korgis

Provides the reusable local AI execution boundary beneath the desktop workflow.

Local model runtime

Runs the GGUF inference path on infrastructure controlled by the user.

Explore the architecture on GitHub ↗
⚠

RedactGuard is an experimental privacy tool, not a compliance guarantee. Local AI can miss, misclassify or over-detect sensitive information, so results still require human review.

EVIDENCE & LIMITS

The desktop workflow is implemented. Detection is still fallible.

RedactGuard treats reviewability as part of the product boundary: local inference can accelerate detection, but it does not turn the model into an autonomous privacy or compliance authority.

EVIDENCE BOUNDARY

Local, configurable and reviewable does not mean perfect.

The current desktop product supports configurable PII definitions, local execution through the Korgis boundary, reviewable findings and explicit export. The model can still miss, misclassify or over-detect sensitive information, so human review remains mandatory.

Supported today

  • Document analysis can run through a local execution boundary
  • Standard and custom PII definitions are configurable
  • Every finding remains visible and reviewable before export
  • The minimized result moves only after explicit user choices

Not claimed here

  • Compliance guarantee
  • Published perfect-recall or perfect-precision claim across PII domains

CONNECTED SYSTEM

Where this project fits

BUILD · INFRASTRUCTUREKorgisReusable local inference boundary
PRODUCT · DESKTOPRedactGuard DesktopPrivacy-first proving ground
MEASURE · EVIDENCEPerformance LabRuntime viability on real hardware
RedactGuard · Desktop

Private document workflows. Reusable local execution.

The desktop variant separates the RedactGuard product workflow from the Korgis execution backbone, while keeping the final privacy decision with the user.

Explore Desktop ↗