Changelog

Industry, feature, comparison, use-case, case-study, and blog pages now include a Related Resources section with contextual links to related content — improving navigation and helping visitors discover relevant pages.

Every VS comparison page now links directly to the Arbitex feature pages discussed in the comparison narrative. Visitors can jump from a competitive comparison to the specific capability details without returning to the main navigation.

Each industry page now links to its corresponding use-case page, relevant feature pages, and applicable blog content — creating a complete content loop for visitors researching a specific vertical.

Technical blog posts now include a Related section at the end of each article with links to the product features and industry pages discussed in the post. Readers can move from educational content directly to product capabilities.

Twenty-nine blog posts had meta descriptions outside the 150–160 character SEO target range. All have been rewritten to fit within the recommended length while preserving the original meaning and search intent.

The ROI page now includes an interactive calculator that estimates cost savings, risk reduction, and compliance time saved based on your AI usage profile — monthly API calls, average tokens per call, and estimated sensitive data percentage. All calculations run client-side.

A new case study covers how a federal civilian agency deployed Arbitex Outpost in air-gap mode across 14 bureaus — protecting CUI, meeting FISMA control requirements, and providing tamper-proof audit evidence for Inspector General examination.

The policy engine now includes pre-built rule templates for Controlled Unclassified Information — covering CUI markings, contract identifiers, personnel data, and budget context detection across all three DLP tiers.

Audit log exports now support structured formats aligned with NIST SP 800-53 AU-family controls. Export packages include per-user attribution, enforcement action summaries, and audit chain verification evidence suitable for IG and GAO review.

Outpost deployments now support bureau-level or department-level policy overrides within a single agency configuration. Administrators can set base rules centrally and allow individual units to activate additional detection rules where mission requirements differ.

The vendor cost comparison slider on the ROI page now correctly announces value changes to screen readers. The aria-valuenow attribute updates in real time as the slider moves, and the vendor count label uses proper singular/plural forms.

A new technical blog post walks through the 3-tier DLP pipeline architecture — how pattern matching, named entity recognition, and contextual validation work together, when each tier fires, and how short-circuit cascades keep latency low while maintaining detection accuracy.

A new blog post covers the policy engine in depth — XACML-inspired combining algorithms, multi-condition rules, compliance bundles, simulation mode, and the admin portal workflow for building and testing policy changes.

Per-entity confidence scores now carry through the full pipeline — from initial entity detection through Tier 3 contextual validation. Audit log entries include the composite score at each stage, giving compliance teams a complete picture of why each entity was flagged or dismissed.

Simulation mode reports now include per-rule impact summaries showing how many requests each rule would affect. The before/after comparison highlights newly blocked, newly allowed, and unchanged request counts — making it easier to assess policy changes before promoting to production.