All Case Studies
CASE STUDY
Automated Book Generation Platform
Client details redacted for privacy. Results verified.
INDUSTRY Genealogy
TIMELINE 8 months
TEAM SIZE 2 engineers
STACK AWS
The Challenge
Print-Ready PDF Generation at Scale
A genealogy-focused client needed a platform to automatically compile family history data
into professionally printed books. Users import records from Ancestry.com, and the system
generates real-time previews and high-fidelity, print-ready PDFs. Output sizes range from
tens to hundreds of megabytes per book.
Requirements: - Real-time preview generation during the editing process
- Print-ready PDF rendering at production quality
- 20,000+ monthly active users, many spending hours daily
- Zero data loss across the pipeline
- High-volume batch processing for preview and PDF data
The Solution
Cloud-Native Microservices on AWS
We built a cloud-native microservices architecture on AWS, with compute distributed across
ECS for persistent services and AWS Batch paired with Lambda for heavy processing workloads.
The entire infrastructure was defined as code from the start.
Core Stack ECS (Fargate) Microservices for book compilation, data ingestion, and user session management
AWS Batch Large-scale batch generation of preview and PDF assets
Lambda Event-driven processing for import triggers, image transformations, and notification dispatch
S3 Storage for imported data, generated previews, and final PDF output
End-to-end delivery - Ancestry.com data import and normalization pipeline
- Real-time preview rendering engine
- High-fidelity PDF generation for commercial print
- Sustained concurrency for thousands of active users with isolated processing queues
Resiliency
Multi-Region Failover With Zero Data Loss
A single-region deployment was not acceptable for a product handling irreplaceable family
records. We configured multi-region failover across us-east-1 and us-west-2 to guarantee
high availability.
- 1.
Active-passive failover with automated health checks
- 2.
Cross-region replication for all stored assets and user data
- 3.
DNS-based routing with sub-minute failover detection
No user-facing downtime occurred during the project's production lifecycle.
Results
20,000+ Monthly Users, Zero Downtime
The platform processes family history imports, generates real-time previews, and delivers
print-ready PDFs at scale — all running on a fully automated, resilient AWS backbone.
Metric
Value
Monthly active users 20,000+
PDF output size Tens to hundreds of MBs
Regional failover us-east / us-west
Data loss incidents 0
Team size 2 engineers
Technologies
AWS ECS Batch Lambda S3 Multi-Region IaC