Global Air Cargo Knowledge Management System

Global Air Cargo Knowledge Management System

Challenge GroupIsrael - Malta
https://www.challenge-group.com

Challenge Group, the global leader of door 2 door logistics of complex air cargo, operates a self-sufficient ecosystem across airfreight, handling & logistics, and maintenance & aviation services. Critical operational knowledge lived in tens of thousands of documents and systems—flight schedules, handling procedures, maintenance logs, product SOPs (C-VAL, C-PHARMA, C-FRESH, C-ECOM, C-DGR, C-CAR, C-BIG, C-AVI), and regulatory manuals—spread across multiple platforms. Teams on the ground, in the air, and at HQ struggled to get a unified, up-to-date picture fast enough to support high-stakes decisions.

Challenge Group

Industry

Aviation & Air Cargo Logistics

Timeline

8 months

Team Size

12 specialists

The Challenge

Challenge Group, the global leader of door 2 door logistics of complex air cargo, operates a self-sufficient ecosystem across airfreight, handling & logistics, and maintenance & aviation services. Critical operational knowledge lived in tens of thousands of documents and systems—flight schedules, handling procedures, maintenance logs, product SOPs (C-VAL, C-PHARMA, C-FRESH, C-ECOM, C-DGR, C-CAR, C-BIG, C-AVI), and regulatory manuals—spread across multiple platforms. Teams on the ground, in the air, and at HQ struggled to get a unified, up-to-date picture fast enough to support high-stakes decisions.

Key Pain Points:

  • Operational knowledge spread across flight schedule tools, handling and warehouse systems, maintenance systems, and document repositories

  • No unified search across internal manuals, product SOPs, and regulatory documents—employees had to know which system to query

  • Duplicate and conflicting procedures for similar lanes, products, or stations

  • Outdated process documents mixed with current best practices for air cargo operations

  • Knowledge silos between airfreight, handling & logistics, and maintenance & aviation teams

  • Frontline and junior staff couldn’t easily access the expertise of senior operations and product specialists

  • Onboarding new employees at stations and offices took many months before they could confidently navigate the knowledge landscape

Business Impact

The fragmented knowledge landscape meant slower responses to operational questions, longer time to prepare complex shipments, and increased risk of rework or delays. Productivity losses ran into the tens of millions SEK annually when aggregated across global stations, with additional opportunity costs from slower solution design for customers with complex cargo needs.

Our Solution

We built a custom RAG (Retrieval-Augmented Generation) system that unified Challenge Group’s operational and product knowledge into a single intelligent search interface. The system understands logistics context, retrieves relevant information across all sources, and provides answers in natural language for teams working in AirFreight, Handling & Logistics Services, and Maintenance & Aviation Services.

Our Approach

1
Discovery & Data Audit (Month 1-2)

Catalogued all knowledge sources across airfreight, handling & logistics, and maintenance & aviation, assessed data quality, interviewed 200+ employees (from station handlers to central operations leaders) about information needs, and mapped end-to-end knowledge workflows for complex air cargo.

2
RAG Architecture Design (Month 2-3)

Designed a vector database schema tailored to air cargo operations, selected embedding models, built a retrieval pipeline that understands product codes (C-VAL, C-PHARMA, C-FRESH, C-ECOM, C-DGR, C-CAR, C-BIG, C-AVI) and operational terminology, and developed prompt engineering strategies for accurate, context-aware responses.

3
Data Processing & Ingestion (Month 3-5)

Built ETL pipelines for each data source—document repositories, SOP libraries, knowledge bases, and selected operational data—cleaned and standardized content, created embeddings for 50K+ documents, and implemented an incremental update system that keeps new procedures and regulatory changes in sync.

4
Interface Development (Month 5-6)

Built a web-based chat interface for knowledge queries, developed a Slack integration for teams working in real time, created a browser extension for in-context search in existing tools, and built an admin dashboard for monitoring usage, feedback, and content health.

5
Testing & Refinement (Month 6-7)

Ran beta testing with 500 employees across multiple stations and functions, refined retrieval accuracy on complex cargo scenarios, improved response quality with domain-specific prompts, optimized performance, and trained champion users in operations and product teams.

6
Rollout & Training (Month 7-8)

Executed a phased rollout across regions and business units, delivered tailored training for frontline, operations, and support teams, provided comprehensive documentation, and set up an ongoing feedback loop with Challenge Group’s digital and operations leads.

Technical Highlights

Custom RAG architecture with Pinecone vector database optimized for air cargo and logistics knowledge

OpenAI GPT-4 for response generation with domain-tuned prompting

Custom fine-tuned embedding models for aviation and air cargo terminology

Hybrid search: vector similarity + traditional keyword matching to handle codes, AWB numbers, and free text

Real-time ingestion from multiple knowledge sources and document systems

Permission-aware retrieval (users only see the procedures and documents they have access to)

Source attribution and citation for all responses to ensure traceability and compliance

Feedback loop for continuous improvement driven by frontline usage

Multi-language support (English, French, German, Spanish) to reflect Challenge Group’s global footprint

The Results

The system fundamentally changed how Challenge Group teams access and use knowledge across their air cargo and logistics operations.

-70%

Search Time Reduction

From an average of 2.5 hours to 45 minutes daily spent searching for procedures and answers

92%

Information Accuracy

Correct operational answer on first attempt in internal evaluations

+45%

Employee Satisfaction

Improvement in information access satisfaction scores across stations and HQ

94%

Adoption Rate

Active users within 3 months across airfreight, handling & logistics, and maintenance & aviation teams

81.6M SEK

Productivity Savings

Annual time savings valued across the organization

-40%

Proposal Development Time

Faster access to past solutions and product SOPs for complex cargo proposals

-50%

Onboarding Time

New employees becoming productive faster thanks to a single source of truth

+80%

Knowledge Retention

Critical operational and product knowledge now captured and searchable

Business Impact

01

Accelerated preparation and validation of complex air cargo solutions across products like C-VAL, C-PHARMA, and C-FRESH

02

Reduced dependency on a small group of senior experts for routine knowledge and procedural questions

03

Improved cross-department collaboration between airfreight, handling & logistics, and maintenance & aviation teams

04

Enhanced compliance and audit readiness through easy access to the latest procedures and policies

05

Reduced employee frustration and improved retention by making knowledge access fast and reliable

Return on Investment

Investment:

3,400,000 SEK

Annual productivity savings:

81.6M SEK
ROI:

12 months to break even

Return in first year:

2,353% return in first year

Investment → Revenue Growth

Head of Digital Transformation

"For a group like ours, operating door 2 door logistics of complex air cargo worldwide, having instant access to accurate, up-to-date knowledge is critical. This system lets our teams find clear, actionable answers in seconds instead of digging through multiple systems and documents. It feels like having a senior operations expert available 24/7 for every station."

Head of Digital Transformation

Challenge Group

Project Gallery

Technologies Used

FastAPI
FastAPI
React
React
TypeScript
TypeScript
Pinecone
Pinecone
OpenAI GPT
OpenAI GPT
LangChain
LangChain
PostgreSQL
PostgreSQL
Redis
Redis
Elasticsearch
Elasticsearch
Docker
Docker
Kubernetes
Kubernetes
Azure
Azure

Related Services

Services we used for this project

AI Engineering

Intelligent systems powered by cutting-edge AI

Learn More
Development

Custom software development solutions built to scale

Learn More

Ready to Achieve Similar Results?

Let's discuss how we can help you transform your business with our proven expertise.