Platinur

We're building something cool.

Analytics assets,
shipped trusted

Platinur is a customer-hosted control plane that turns raw ClickHouse data into governed dbt models, Evidence dashboards, GitHub promotions, scheduled refreshes, and monitored production runs.

Start an Assisted Pilot See the Product Flow arrow_forward
Product Loop

From raw sources to trusted production analytics.

Platinur gives operators one governed loop for source onboarding, AI-assisted modeling, staging validation, GitHub promotion, production refreshes, and failure repair. Every generated change becomes a traceable artifact instead of an unchecked dashboard edit.

SOURCES

One schema per source

Connect sources or use existing ClickHouse schemas, test the connection, select all tables or specific tables, and make the source contract explicit before modeling begins.

ClickHouse Integrations Schemas Tables
STAGING

Fast governed development

Initial Run and assistant changes generate dbt models and Evidence dashboards in staging, validate them, run selected models, and preview before production release.

Initial Run Assistant dbt Evidence
PROD

Release and monitor

Promotion PRs move validated staging changes to main. Production runs from governed code, schedules refreshes, and records each run and audit event.

GitHub PR Prod branch Schedules Monitoring
ASSISTED CONTROL
Validated proposals
Model explanations
Repair journeys
Release tracking
What Platinur Does

The control plane for analytics release.

verified_user

Sources and Initial Run

Configure sources, test connections, select schemas and tables, add modeling context, then generate the first dbt and Evidence project from the source contract.

Start pilot setup
monitoring

Staging development

The assistant creates validated proposals in staging, applies them to the staging branch, runs changed models, and builds staging dashboard previews before release.

Assistant / dbt / Evidence / Staging

database

Promotion and monitoring

Promote staging to production through GitHub, run production refreshes from governed code, and inspect every run, failure, audit event, and repair path.

GitHub / Prod / Schedules / Runs

Customer-hosted by design

Runs on the customer stack with ClickHouse, dbt Core, Evidence, Prefect, GitHub, and a bounded OpenAI-backed assistant.

auto_awesome
Selected Capabilities

A productized stack for governed analytics delivery.

Warehouse

ClickHouse

Transformation

dbt Core

Dashboards

Evidence

Orchestration

Prefect

Governance

GitHub

Assistant

OpenAI

Example pilot

Raw data to governed production analytics

"Our source data is in ClickHouse, but we need a governed way to build models, dashboards, and production releases."

Input state

  • ClickHouse contains source schemas
  • Teams need selected tables modeled into a DW layer
  • Dashboards need preview before release
  • Production changes need traceability

Platinur path

  • Select sources and run Initial Run
  • Validate and apply proposals in staging
  • Promote via GitHub into production
  • Monitor runs, schedules, and failures
How It Works

A governed loop, not a chatbot bolted to BI.

Platinur makes each step explicit: generate a candidate, validate it, apply it to staging, run and preview, promote with GitHub, refresh production, then monitor and repair from the same control plane.

  • check_circle

    1. Define sources

    Register source schemas and choose all tables or selected tables for the analytics build.

  • check_circle

    2. Generate and validate

    Use Initial Run or the assistant to create proposals with dbt and Evidence validation artifacts.

  • check_circle

    3. Prove in staging

    Apply the proposal to staging, run changed models, and open a staging dashboard preview.

  • check_circle

    4. Promote to prod

    Create the staging-to-main GitHub PR and let production run from the governed main branch.

  • check_circle

    5. Monitor and repair

    Track runs, audit events, dashboard freshness, and failure repair from Monitoring.

The assistant works inside product rules.

Development happens in staging, proposals must validate, and production remains governed by the release path.

Explore the product