MATIH Platform is in active MVP development. Documentation reflects current implementation status.
2. Architecture
Overview

Service Topology Overview

The MATIH Platform comprises 24 microservices distributed across two planes and seven Kubernetes namespaces. This section maps the complete service topology, including service discovery, dependency graphs, communication patterns, and failure propagation analysis.


Topology Layers

LayerServicesCommunication
External EdgeBrowser, CLI clientsHTTPS to Kong Gateway
Control Plane Mesh10 Java/Spring Boot servicesREST + Kafka
Data Plane Mesh14 polyglot servicesREST + Kafka + Redis Pub/Sub

Sections

SectionDescription
Service RegistryService discovery, DNS names, and configuration
Communication PatternsREST, Kafka, Redis Pub/Sub, gRPC patterns
Dependency MapFull dependency graph and failure analysis

Service Count

MetricValue
Total microservices24
Control Plane services10 (all Java/Spring Boot 3.2)
Data Plane services14 (Java, Python, Node.js)
Frontend applications8 (React/Vite)
Kubernetes namespaces7
Data store technologies9

Critical Path

The following services are on the critical path for the "Intent to Insights" workflow:

api-gateway --> iam-service --> ai-service --> query-engine --> Trino
                                    |
                                    +--> semantic-layer
                                    +--> catalog-service

A failure in any of these services disrupts the primary user experience. These services are deployed with 3 replicas in production.


Related Sections