First enterprise release — May 29, 2026
This is the first version of Verify Freight released for enterprise use. This platform provides an automated, highly secure compliance pipeline for hazardous materials shipping manifests. By combining computer vision, a deterministic rules engine, and focused AI reasoning, the system instantly validates shipments against Title 49 of the Code of Federal Regulations (CFR) that contains the hazardous materials regulations (HMR) issued by the Pipeline and Hazardous Materials Safety Administration (PHMSA) — Parts 172 and 177.
Users can drag and drop manifests in PDF, CSV, or image formats — including scanned documents and spreadsheets.
AWS Textract automatically extracts key structured data — UN numbers, hazard classes, proper shipping names, packing groups, and quantities.
Standard validations are executed instantly — with absolute certainty and without AI — by cross-referencing extracted fields against the 49 CFR §172.101 Hazardous Materials Table.
AI is reserved strictly for complex regulatory checks — load compatibility (§177.848), placarding thresholds (§172.504), and special provisions (§172.102). This targeted approach keeps processing highly efficient and accurate.
The system generates a professional PDF detailing regulatory findings by severity — Critical, Warning, and Advisory — including specific CFR citations and recommended corrective actions.
Every finding clearly indicates whether it was generated by the Rules Engine or AI, ensuring full transparency for DOT audits.
Reports are saved to Amazon S3, triggering an automated email to the user with a secure, 7-day download link.
Manifests, reports, and job histories are completely segregated to ensure strict data privacy between different customer environments.
Protected by AWS WAF (OWASP rules), Cognito JWT authentication for APIs, customer-managed KMS encryption at rest, and TLS in transit. All S3 access is tracked via CloudTrail audit logging.
Users maintain access to a fully searchable job history dashboard featuring status badges and violation counts. Records are archived for 90 days in development environments and 3 years in production.
For any questions, issues, or feedback regarding this release, please reach out to our team: