MATIH Platform is in active MVP development. Documentation reflects current implementation status.
1. Introduction
Getting Started Guide

Getting Started Guide

This guide provides role-based reading paths through the MATIH Platform documentation. Based on your role and goals, follow the recommended path to build your understanding efficiently. Each path is ordered to build knowledge progressively.


Choose Your Path

If You AreStart HereThen ReadGoal
Evaluating the platformVision and MissionProblem Space, CapabilitiesUnderstand the value proposition
Data EngineerTechnology StackArchitecture, Data Stores, PipelinesPlan integration
BI DeveloperCapabilities (BI)AI Service, BI Service, DashboardsBuild dashboards
ML EngineerML InfrastructureML Service, Ray, MLflowTrain and deploy models
Platform AdministratorArchitecture OverviewKubernetes, Helm, ObservabilityDeploy and operate
Security ReviewerTenant IsolationJWT Tokens, RBAC, Network PoliciesAudit security posture
Frontend DeveloperFrontend StackWorkbench UI, API ClientBuild UI features

Path 1: Platform Evaluator

If you are assessing whether MATIH fits your organization's needs, follow this reading order:

  1. Vision and Mission -- Understand the founding principles
  2. Problem Space -- See the problems MATIH solves
  3. Platform Capabilities -- Review the six capability pillars
  4. User Personas -- Match your team to platform personas
  5. Architecture Preview -- High-level system structure
  6. Design Decisions -- Understand trade-offs

Estimated reading time: 2-3 hours


Path 2: Backend Developer

If you are building or maintaining backend services:

  1. Technology Stack: Backend -- Languages and frameworks
  2. Architecture Deep Dive -- Full system architecture
  3. Control Plane or Data Plane -- Your service area
  4. API Design -- REST conventions and error handling
  5. Multi-Tenancy Architecture -- Tenant context propagation
  6. Event-Driven Architecture -- Kafka and event patterns
  7. Security -- Authentication, authorization, and isolation

Estimated reading time: 4-6 hours


Path 3: Data Engineer

If you are integrating data pipelines or connecting data sources:

  1. Technology Stack: Data Infrastructure -- Data technologies
  2. Data Stores -- PostgreSQL, Kafka, Trino, and more
  3. Data Flow -- Request lifecycle and query flow
  4. Pipeline Flow -- Pipeline orchestration
  5. Compute Engines -- Trino, Spark, Flink

Estimated reading time: 3-4 hours


Path 4: ML Engineer

If you are training models or deploying ML workflows:

  1. Technology Stack: ML Infrastructure -- ML/AI technologies
  2. ML Flow -- Model training and serving lifecycle
  3. Compute Engines -- Ray, Spark
  4. Data Stores: Vector Stores -- Qdrant, LanceDB

Estimated reading time: 2-3 hours


Path 5: Platform Administrator

If you are deploying, configuring, or operating the platform:

  1. Architecture Preview -- System structure
  2. Service Topology -- Service dependencies and failure domains
  3. Multi-Tenancy -- Namespace and resource isolation
  4. Technology Stack: Orchestration -- Kubernetes, Helm, Terraform
  5. Security -- Full security model

Estimated reading time: 4-5 hours


Path 6: Security Reviewer

If you are auditing the platform's security posture:

  1. Security Overview -- Full security model
  2. Tenant Isolation -- Multi-layer isolation
  3. JWT Tokens -- Token structure and lifecycle
  4. RBAC -- Role hierarchy and permissions
  5. Multi-Tenancy: Database Isolation -- Data isolation
  6. Multi-Tenancy: Namespace Isolation -- Infrastructure isolation
  7. Encryption -- Data at rest and in transit

Estimated reading time: 3-4 hours


Quick Reference

Key Terminology

Before diving deep, familiarize yourself with the Key Terminology page. It defines 100+ terms organized by domain, covering the Control Plane, Data Plane, AI agents, multi-tenancy, and infrastructure concepts.

Documentation Conventions

ConventionMeaning
monospace textCode, commands, file paths, service names
Bold textTerms being defined or important notes
Tables with "Planned" markersFeatures designed but not yet implemented
Port numbers in parenthesesDefault service port, e.g., ai-service (8000)

Getting Help

ResourceDescription
Architecture OverviewComplete system architecture reference
Service TopologyService dependency and failure analysis
Key TerminologyGlossary of platform terms

Next Steps

Choose the path that matches your role and begin reading. Each chapter builds on the concepts introduced in this introductory chapter. If you have not yet read the Vision and Mission section, it provides essential context for understanding the design decisions throughout the platform.