Kintex Hybrid Analytics Architecture

Three clear phases to modernize Kintex analytics: run a nightly ETL, explore metrics with Metabase, and deliver a focused dashboard.

Separate Data Store Faster Metrics One Dashboard

Guiding Objectives

  • Keep Kintex reporting data in its own store so the EPB ERP stays steady.
  • Let teams test KPIs quickly in Metabase before any code work.
  • Show long-term trends and live KPIs in one branded dashboard.
Overview

Why the Kintex hybrid rollout happens in phases

The hybrid architecture is intentionally sequenced so each stage delivers value on its own and lays the groundwork for the next. By finishing the extract-and-load work first, teams can experiment in Metabase with real data, then promote proven insights into a branded Kintex experience without rework.

Phase 1 · Establish the analytics store

Stabilize a nightly ETL from EPB so reporting data lives outside the ERP and history is preserved.

Phase 2 · Validate metrics with Metabase

Use low-code exploration to agree on KPIs, iterate quickly, and document the definitions Kintex cares about.

Phase 3 · Launch the Kintex experience

Embed the curated content and custom components to give every team a polished, always-on analytics hub.

Three-Phase Plan

Build the pipeline step by step

Finish one stage, learn from it, then move on with the pieces already in place.

graph TD SRC[EPB ERP Database] -->|Nightly Extract| CLEAN[Transform & Clean] CLEAN --> AGG[Aggregate & Model] AGG --> DEST[Kintex Analytics PostgreSQL]

Diagram: Nightly ETL flow from EPB to the Kintex analytics store.

Phase 1 · Nightly ETL

Copy EPB ERP data into a Kintex analytics store every night, shaping it for reporting and keeping history intact.

What happens

  • Nightly extract and transform jobs
  • Cleaned tables built for analysis
  • Historical snapshots retained

Why it helps

  • ERP stays focused on transactions
  • Analytics queries run faster
  • Data foundation ready for new Kintex use cases
Metabase dashboard with sales metrics

Screenshot: Metabase dashboard sample.

Phase 2 · Metabase Exploration

Use Metabase to explore the analytics tables, test KPIs with Kintex stakeholders, and agree on the metrics that matter.

What happens

  • Teams build charts without code
  • Dashboards shared for feedback
  • Metric definitions captured

Why it helps

  • Faster iteration on ideas
  • Less engineering rework
  • Shared understanding of Kintex KPIs
Custom Kintex analytics dashboard concept

Screenshot: Custom Kintex dashboard concept.

Phase 3 · Custom Experience

Embed the finalized Metabase content and add custom widgets for live KPIs, Kintex branding, and advanced filters.

What happens

  • Embed approved Metabase charts
  • Add custom widgets for live KPIs
  • Apply Kintex branding and layout rules

Why it helps

  • Single pane for Kintex insights
  • Supports advanced filters and actions
  • Ready for secure embedding upgrades
Architecture Highlights

Data flow at a glance

See how data moves from the EPB ERP into tools that Kintex teams use every day.

EPB Operational DB

Live system that holds sales, inventory, and customer data.

Nightly ETL

Jobs pull, clean, and prepare the data every night.

Kintex Analytics PostgreSQL

Reporting tables ready for queries and dashboards.

Metabase + Kintex UI

Metabase, APIs, and custom pages show the right insight to each team.

Visual Pipeline

Flow from source to dashboard

Follow the path from the EPB ERP database through ETL, Metabase, and Kintex's custom features.

graph LR A[EPB Operational Database] -->|Nightly Sync| B[ETL Pipeline] B -->|Transformed Models| C[Kintex Analytics PostgreSQL] C -->|Direct Connection| D[Metabase Exploration] C -->|APIs & Views| E[Kintex Dashboard Services] D -->|Embedded Charts| F[Kintex Dashboard] E -->|Real-time Widgets| F D -->|Validated KPIs| G[Kintex Metric Catalog] G --> F

Why the order matters

  • ERP stays fast because reporting runs on a different database.
  • Metabase helps pick the right KPIs before we build custom code.
  • The final dashboard blends long-term trends and live widgets in one place.

Implementation Priorities

Data Readiness

Pick the ETL tool (Laravel Scheduler, Airflow, or cron) and shape analytics tables.

Metric Catalog

Write down KPI definitions as Metabase sessions confirm them with stakeholders.

Experience Layer

Design the dashboard layout that mixes embedded charts and custom widgets.

Key deliverables

ETL Playbook

Nightly sync scripts, job steps, and basic monitoring tips.

Metabase Prototype

Sample dashboards showing agreed KPIs and notes from reviews.

Custom Dashboard Shell

Responsive page ready for embeds and custom widgets.

Adoption Toolkit

Short guides, walkthroughs, and a simple governance checklist.