Dangerous goods validation, automated
30 min → 50 sec
AI bot reads shipper emails and PDFs, cross-checks every booking against the live IMDG knowledge base — 50 seconds end-to-end.
Anonymised customer profile
Industry
Container Line
Region
Middle East
Fleet / scale
200+ daily DG requests across 8 specialists
The challenge
30% of DG booking requests were being lost, delayed beyond cutoff, or abandoned in manual processing before a specialist ever reviewed them.
- Compliance officers manually cross-referenced safety data sheets against IMDG, SOLAS and MARPOL — 15 to 45 minutes per booking.
- At 200+ daily requests across 8 specialists, queues stretched to 24 hours; bookings missed vessel cutoffs.
- Under time pressure near cutoffs, teams defaulted to rejection — losing compliant cargo that could legally have been accepted.
- Audit trail scattered across email, spreadsheets, ERP records and shared folders.
The approach
Deployed ASTRA DG Bot — an intelligence layer between booking input channels and the carrier's ERP, with expert decision control retained.
- Multi-channel ingestion: emails (PDF/Excel/Word/text), web portals, API feeds, EDI messages.
- AI data extraction in 20 seconds — UN numbers, hazard classes, quantities, flash points, marine-pollutant status.
- Parallel validation in 30 seconds — IMDG Code, SOLAS stowage, port restrictions, carrier-specific rules.
- Expert makes the final call; full digital audit trail attached.
The results
30 min → 50 sec
Per-booking validation time
30% → 0%
DG bookings lost in manual queues
80% → 100%
Expert time spent on decisions, not data extraction
70-85% → 100%
Audit trail completeness
Deployment
4 weeks integration · 4 weeks parallel validation · then live
ROI
Phased rollout — expert-supervised processing from week 9
Talk to us about a similar deployment.
We'll map your workflow, share what we've learned from similar engagements, and tell you honestly where the gains sit.
Book a 30-min conversation