Analytics  ·  intelligence

Run on data.
Powered by proof.

Dashboards your team will trust. KPIs everyone agrees on. Data refreshed before you ask for it. Built on open source you own and we maintain.

Metabase. Superset. dbt. Postgres. Yours forever.

A live preview

One chart. Four tiles. Live.

A SaaS-startup operating dashboard the morning after we ship. Revenue trend on top, KPI tiles with sparklines below. Sage = good, brass = needs attention.

Monthly recurring revenue · last 12 months

R 1 842 000

5.6%vs prior month
R782kR1070kR1358kR1646kR1934kJUNJULAUGSEPOCTNOVDECJANFEBMARAPRMAYR1842k
Live
1 247
Active customers
+82 / 30d
Live
21d
Avg deal cycle
-3d / 30d
Live
2.4%
Churn (30d)
-0.6pp / 30d
Live
R 840
Customer acquisition cost
+R 120 / 30d

Demo data · your dashboards run on your numbers · refreshed every 15 min

73%

Of business decisions made on gut feel

Gartner, 2024

60%

Of analyst time spent finding and cleaning data

IBM, 2024

More likely to make decisions faster, with data

McKinsey, 2024

R47M

Average revenue uplift, data-driven SA companies

PwC SA, 2024

What we build

Six things every business should measure.

DASHBOARDS

Live Dashboards

Metabase or Superset, connected to your warehouse. Daily refresh, hourly refresh, real-time if you need it. Your team sees the same number.

WAREHOUSE

Data Warehousing

PostgreSQL warehouse modelled around how your business thinks. Sources unified, cleaned, version-controlled.

PIPELINES

ETL & Pipelines

Airbyte, dbt, Airflow. CRM, accounting, ops, marketing — piped to the warehouse, transformed into something useful.

DEFINITIONS

KPI Definitions

Six different definitions of 'active customer' across departments? We define each metric once. Everyone sees the same number.

REPORTING

Automated Reporting

Monthly board pack. Weekly ops dashboard. Daily alerts. All running themselves on a schedule. Tuesdays free up.

SELF-SERVE

Self-Serve Analytics

Your team writes their own queries with guard-rails. SQL training included. The ad-hoc requests stop piling up on one person.

Real wins

Before and after the warehouse.

Six KPI definitions, one source of truth

Before

Sales, marketing, and finance each had their own 'active customer' number. Reports disagreed by 18%. Every meeting started with a debate about the data.

After

Metric layer in dbt with definitions owned by each team-lead. One number, agreed in writing. Meetings start from the data.

dbt · PostgreSQL · Metabase

P&L from 3 weeks late to next-day

Before

CFO got the monthly P&L pack on the 18th of the following month. By the time it landed, decisions had already been made on gut feel.

After

P&L runs nightly. Drill-downs by GL account, cost centre, project. CFO checks it with morning coffee on the 1st.

Airbyte · dbt · Superset · Slack alerts

Marketing's weekly deck builds itself

Before

Manager spent every Monday morning copying numbers from Google Analytics, Meta, Mailchimp, the CRM into a PowerPoint. Half a day, every week.

After

Live dashboard pulling all four sources. Slides auto-generate from the dashboard as a Monday-morning email. Manager got half a day back.

Airbyte · ClickHouse · Metabase · n8n

The stack

What's under the dashboards.

WarehousesPostgreSQL · ClickHouse · DuckDB · BigQuery · Snowflake. Self-hosted preferred, cloud where it makes sense.
Modellingdbt · SQLMesh. One metric layer, version-controlled, code-reviewed like the rest of your codebase.
PipelinesAirbyte · Fivetran · custom Python. Whatever moves your data cleanest, scheduled by Airflow or Dagster.
VisualisationMetabase · Apache Superset · Cube · Grafana. Open-source first. Your dashboards, your infra, your licence-free.
Real-timeMaterialize · ClickHouse · Kafka. When 'fresh by tomorrow' isn't fresh enough.
IntegrationsWebhooks to Slack, Teams, WhatsApp. Auto-generated PDFs. Scheduled emails. The dashboard tells the team without anyone asking.

How an engagement runs

Five stages, audit to live.

01

Audit

Map what you measure today and what you should be. We find the metrics nobody trusts, the reports nobody reads, the questions nobody can answer.

02

Model

Define every metric and entity once, agreed across teams. Locked in the metric layer so nobody has to reinvent 'monthly revenue' next quarter.

03

Pipe

Connect sources, build the warehouse, schedule the refreshes. Audit trail on every load so you know when the data last updated.

04

Visualise

Dashboards in Metabase or Superset. Tuned for the question the team is trying to answer.

05

Train

Your team learns the metric layer, writes their own queries with guard-rails. The on-call analyst stops being everyone's bottleneck.

Tell us what you measure.
We'll build the dashboard.

Free 30-minute call. We map the metrics, the sources, and what a real dashboard would look like for your team. Quote follows after.