5.7

Release Date: April 20, 2026

We’ve implemented several key improvements to the processing engine to enhance the handling of formerly failed files and to resolve intermittent processing stalls:

• Resolved Processing Failures: The engine now successfully extracts and processes complex .docx, .pptx, and .msg files, as well as large .txt files.
• Skipped .p7s file: We’ve refactored the engine to automatically skip .p7s files. Since these files typically do not contain extractable PII, skipping them reduces processing noise and improves overall speed.
• Enhanced Pipeline Resilience: We’ve addressed an issue where certain “unknown” file types caused processing stalls or pipeline crashes. The system now handles these unknown file types gracefully, ensuring an uninterrupted processing flow.

We’ve implemented refined algorithms to optimize our detection models, providing more accurate identification of Canadian Social Insurance Number (SIN) and reducing false positives. This ensures high precision during review and supports better compliance with Canadian data privacy regulations.

We’ve resolved an issue that prevented users from downloading files that failed during Processing, specifically when those files are large or exceed certain size thresholds. Users can now consistently download these large failed files.

Instead of an immediate download, the system will generate a secure download link once the file is ready. This update ensures that even the largest failed files can be downloaded for local troubleshooting or re-processing.