Sky-High Insights: AWS Data Mesh That Unified 100 Databases for Airline Excellence

Slides Deck for the presentation I gave at AWS Community Day Adria (Sarajevo in September 2025.) and AWS Community Day Bulgaria (Sofia in October 2025.).
Speakers Lists:
https://awscommunityadria.com/2025/speakers/
https://www.aws-community-day.bg/speakers-2025

Transcript

1. Sky-High Insights: AWS Data Mesh That Unified 100 Databases for Airline Excellence Bojan Sovilj 25.10.2025.
2. CERTIFIED CERTIFIED
3. Agenda: Enterprise data challenges What is Data Mesh (Airline use case) Key terminology and concepts Core AWS principles for scaling Data analytics workflow • Ingestion • Transformation • Reporting Complexity in a slide Key results
4. Enterprise Data Challenges
5. Entreprise Data Challenges Central Team Bottleneck - Central team queue: 6-8 weeks average - 40+ pending data requests in backlog - Business decisions delayed Data Quality Issues - Inconsistent definitions across teams - Manual data reconciliation - 30% of reports needs corrections Scaling problems - Central team could’t scale with growth - Domain expertise lost in translation - Tribal knowledge, no documentation Cost Impact - Redundant dana processing - Multiple copies of same data - Inefficient resource utilization Key Metric: Average time from data request to delivery was 2 months
6. revenue accounting aircraft operations airline planning customer services ground operations central data team Obsolete - Central Data Team New - Decentralized Data Ownership revenue accounting aircraft operations airline planning customer services ground operations Data Mesh Data Mesh
7. Domain Ownership Data Product Federated Governance Self Serve Analytics is created and maintained by resides within operates within governs is governed by create data products on Domain Ownership Data Product Federated Governance Self Serve Analytics domain bounded context domain data teams operational & analytical data contextual mapping governance group policy automation product thinking data product by domain team interoperability interfaces domain agnostic data platform team self serve data platform Data Mesh
8. Principles - Federated access control
9. Principles - Automation through CI-CD
10. Data Workload - Ingestion
11. Data Workload Transformation
12. Data Workload Reporting
13. Complexity in a slide
14. Results Results: 11 domains 186 data bases in catalog 50 data products 100 TB data 70 Redshift users 100 + different reports 103 domain persona …
15. Thank you!