Arkkitehtuuri ja skaalautuvuus

Designed for scale

The Debbie infrastructure is hosted entirely on Google Cloud Platform, providing EU-based data sovereignty. The core architecture uses horizontally scalable compute services split into specialized environments:

• Collection Engine Heavy Duty: Dedicated compute for processing massive portfolio uploads and scheduled batch jobs.
• Collection Engine Tasks: Isolated workers for routine background processing, ensuring UI responsiveness never degrades.
• Collection Engine API: The high-availability layer handling synchronous frontend requests, third-party integrations, and webhooks.
• Identity & Access Management (IAM): A dedicated service purely for authentication and strict role-based access control.

Data and Caching Layer

Rather than relying on a single monolithic database, Debbie splits data concerns into purpose-built layers:

• PostgreSQL Replica Sets: Primary persistence is handled by distributed PostgreSQL clusters. Data is separated logically into Integrations, Collection Engine, Webhooks, and IAM.
• Elasticsearch Sharded Cluster: Enables sub-second full-text search and complex querying across millions of debtor records and case histories.
• Redis Memory Cache: Provides extremely fast session state and intermediate data caching for API requests.
• Cloud Tasks & Pub/Sub: All intensive tasks are offloaded to asynchronous message queues, ensuring the system remains highly responsive under load.