MATIH Platform is in active MVP development. Documentation reflects current implementation status.
1. Introduction
Overview

User Personas

The MATIH Platform is designed to serve six primary personas, each representing a distinct role in a modern data organization. Every persona has a dedicated workbench application tailored to their workflow, and all personas share access to the conversational AI interface as their primary interaction model.


Persona Overview

PersonaPrimary WorkbenchCore ActivityKey Platform Capabilities
Data EngineerData Workbench (3002)Build and maintain data pipelinesPipeline orchestration, data quality, catalog management
BI DeveloperBI Workbench (3000)Create dashboards and reportsDashboard builder, semantic layer, scheduled reports
ML EngineerML Workbench (3001)Train and deploy ML modelsExperiment tracking, model registry, distributed training
Agentic UserAgentic Workbench (3003)Ask questions in natural languageConversational analytics, text-to-SQL, AI agents
Platform AdminControl Plane UI (3004)Manage tenants and infrastructureTenant provisioning, monitoring, configuration
ExecutiveBI Workbench (3000)Consume insights and KPIsPre-built dashboards, natural language queries, alerts

The Conversational Interface

While each persona has a specialized workbench, the conversational AI interface is accessible from every workbench. This means that any user can ask a natural language question and receive a data-driven answer, regardless of their technical proficiency.

Any Workbench
  |
  v
Conversational AI Panel
  |
  +-- "What was revenue last quarter?"
  +-- "Show me the top 10 customers by spend"
  +-- "Create a chart comparing Q1 vs Q2 sales"
  +-- "Why did the pipeline fail last night?"
  |
  v
AI Agent Orchestrator --> SQL Generation --> Query Execution --> Visualization

The conversational interface adapts its behavior based on the persona. A Data Engineer asking about pipeline failures receives technical diagnostics, while an Executive asking about revenue trends receives a summary with a chart.


Persona-Workbench Mapping

Each workbench is a standalone React/TypeScript/Vite application designed for a specific persona's workflow:

WorkbenchPortPrimary PersonaSecondary Personas
BI Workbench3000BI DeveloperExecutive, Agentic User
ML Workbench3001ML EngineerData Engineer
Data Workbench3002Data EngineerBI Developer
Agentic Workbench3003Agentic UserAll personas
Control Plane UI3004Platform Admin--
Data Plane UI3005Platform AdminData Engineer

Cross-Persona Collaboration

The platform enables collaboration across persona boundaries through shared artifacts:

ArtifactCreated ByUsed By
DashboardsBI DeveloperExecutive, Agentic User
Semantic modelsBI DeveloperAI Service, ML Engineer
Data pipelinesData EngineerBI Developer, ML Engineer
ML modelsML EngineerBI Developer (predictions in dashboards)
Data quality rulesData EngineerGovernance, BI Developer
Saved queriesAny personaAll personas

Reading Guide

Each persona page in this section follows a consistent structure:

  1. Role description -- Who this persona is and what they do
  2. Day-in-the-life workflow -- A typical workday using the platform
  3. Key capabilities -- The platform features most relevant to this persona
  4. Workbench overview -- The primary UI and its key screens
  5. Related chapters -- Where to find implementation details

Select your persona from the list below to see your tailored guide: