With AI agents in Dataiku, European Air Transport automated document processing to achieve near real-time reporting that informs planning and critical decision-making. 
80x
more productive; 40–50 hours of manual work reduced to 30 minutes
1000s
of documents processed automatically
Reduced
data latency for daily reports from days to just minutes
The Challenge
European Air Transport (EAT), part of DHL Aviation, must analyze vast volumes of operational data each day: flight movements, cargo loads, delays, weather conditions, ground operations, and more.
Across Ground Ops and Fueling teams, operational data lived inside complex, unstructured 40+ page PDFs that included audit reports, vendor fuel receipts, and more. Analysts manually searched each PDF to find information like chemical substances, emissions details, and daily operational insights to inform planning and critical decision-making.
Because these insights fed directly into Power BI dashboards and daily operational decisions, delays of up to a week to complete reports created blind spots. Compressing the reporting cycle enhanced timely visibility and ensured decisions were based on current, accurate data.
The Vision
EAT needed to create a self-service, agentic AI system to instantly read, interpret, and extract structured data from any operational document. That meant business teams would no longer depend on lengthy manual transcription or engineering bottlenecks in order to unlock trapped data instantly and focus on value-added analysis rather than manual entry.
The Execution
Using Dataiku, The Universal AI Platform™, EAT built Scema, an AI-agent-powered document intelligence platform that autonomously interprets documents and orchestrates its own extraction workflow end-to-end.
EAT now uses Scema as a fully self-service extraction engine that gives teams near-real-time visibility into operations. Analysts no longer sift through long PDFs. Instead, they trigger an agent that extracts, validates, generates necessary calculation formulas, and structures information in minutes.
Teams now use these near-real-time insights in Excel and Power BI to keep emissions dashboards current, accelerate ESG submissions, and react to operational issues the same day they arise. Plus, their 'create once, use forever' model has yielded compounding efficiency gains.
Dataiku provided the entire user-facing application, orchestration, and governance framework. It was the ‘last mile’ that made our internal AI accessible to the business.
- Marcus Neufeldt, BI Analyst, Data, European Air Transport (DHL)
What’s Next? EAT is expanding Scema across new document types including contracts, daily audits, and operational inspections. They’re turning the agent into a reusable intelligence layer that powers faster reporting, richer analytics, and broader automation across DHL.
GET TO KNOW EUROPEAN AIR TRANSPORT (DHL)
INDUSTRY
AVIATION
TECH STACK
Dataiku + Snowflake + DHL Internal AI Gateway
KEY CAPABILITIES
GenAI & Agents, AI Governance