

DG Inspector · Dangerous Goods / Hazmat Shipping Compliance
DG Inspector — Dangerous Goods Compliance Software Powered by Document AI
Point. Scan. Comply.
What it does
Built with
FROM THE INSTRUMENT · REAL PRODUCT, NOT MOCKUPS


DG Inspector is dangerous goods compliance software built by Sagan Labs AI that turns a phone camera into a hazmat inspector. It pairs document AI with computer vision so a user can point at a package or upload a shipping document and get an instant verdict — the UN number, hazard class, packing group, and exactly which requirements are missing. It compresses a slow, manual, regulation-heavy inspection into minutes, live at dginspector.com.
The Problem
Shipping dangerous goods is one of the most error-prone tasks in logistics. Every hazmat package has to be reconciled against dense regulations — the IATA Dangerous Goods Regulations for air, the IMDG Code for sea — and then checked for correct labels, placards, packaging markings, and paperwork like the Shipper's Declaration and Safety Data Sheet. A field inspector may run fifty or more of these a day, under time pressure, sometimes with limited formal DG training.
The tools that exist mostly digitize the manual process rather than remove it: they still expect an expert to type in data and interpret the rulebook. A single missed requirement can ground a shipment, trigger a fine, or worse. The industry needed a way to make correct classification fast and repeatable instead of expert-dependent.
Our Approach
We built DG Inspector as an AI-native inspection pipeline rather than a digital checklist. The "Point and Know" flow classifies a package from a single photo. From there a multi-step engine verifies the documents, detects multiple items in a mixed shipment, checks labeling and packaging compliance, applies special provisions, and produces a clear pass/fail verdict with the specific missing requirements called out.
Document AI is the connective tissue: the system reads invoices, packing lists, warehouse receipts, and SDS files to extract the facts a compliance decision depends on, then cross-references them against a regulatory database built from source texts. Because a wrong answer in this domain is unacceptable, the pipeline is engineered for reproducibility, not improvisation.
Architecture Highlights
- Deterministic, evidence-first inspection. Decisions run at temperature zero with multi-run reconciliation for reproducible verdicts, backed by a cited regulatory database spanning all nine hazard classes across air and sea modes — built from source regulations with an audit gate to prevent hallucinated rulings.
- Async inspection engine. Inspections execute as asynchronous cloud jobs; the client subscribes to a live job document that drives a real-time "processing cockpit," so the operator watches the analysis happen.
- One verdict, every device. A shared, pure-TypeScript verdict core is used by both the web app and the React Native mobile app, so a phone in a warehouse and a desktop in a compliance office compute the identical result.
- Split, resilient deployment. A static marketing site is separated from the server-rendered application, each hosted independently for speed and reliability.
- Multilingual by design. Five languages (English, Portuguese, Spanish, French, German) reflect how global freight actually moves.
Outcomes
DG Inspector's goal is to make hazmat compliance fast, consistent, and defensible — turning an hours-long, expert-gated task into a camera-first one, with a documented audit trail behind every verdict.
- Covers all 9 hazard classes across both Air (IATA DGR) and Sea (IMDG) modes.
- Available in five languages — English, Portuguese, Spanish, French, and German.
- Generates the IATA Shipper's Declaration and runs labeling and packaging compliance checks from the same scan.
Ready to make DG compliance fast and reproducible?
DG Inspector is proof of how Sagan Labs AI ships production document AI in a domain where being wrong is not an option. If your team wrestles with dangerous goods compliance software, manual classification, or Shipper's Declaration paperwork, we build the same rigor — deterministic AI, cited regulations, and a real audit trail — around your process. See how we build for regulated industries or talk to us about your compliance workflow, and browse the rest of our work.
Want AI like DG Inspector built around your operation?
Sagan Labs AI designs and ships production AI systems — computer vision, document AI, agents, and automation — tailored to how your business actually runs.
Request a consultation

