Contact
Rodan Data Stack · Field guide

The Data Maturity Pyramid™

Five levels between locked-away data and a system you can ask questions of. Most organisations sit near the bottom while asking for the top. This is the map we use to show you where you are, and the shortest route up.

Score yourself in ~3 minutes Read the framework
The model

You cannot skip a level.

Each level depends on the one beneath it. Analytics on untrusted data is guesswork. AI on a broken foundation is theatre. Build up, in order.

L5AI-driven insightsCan you ask your data questions and act on the answer?
L4Data analyticsCan you see what is happening, and why?
L3Data integrationIs it in one place and joined up?
L2Data qualityCan you trust what you pull?
L1Data collectionCan you reach your data at all?
Where most sit. Levels 1 to 2, moving data by hand and arguing about which number is right.
Where they want to be. Level 5, talking to their data. The gap between the two is the work.
L5

AI-driven insights

Stuck: the ambition, on foundations that will not carry it.

We deliver natural language over your own data, forecasting and decision support, with privacy-preserving deployment where you are regulated.

Talk to it · Eclipse
L4

Data analytics

Stuck: dashboards no one trusts, reporting done by hand.

We deliver reporting, visualisation and AutoML, so the numbers explain themselves.

Visualise · AutoML
L3

Data integration

Stuck: ten tools, ten versions of the truth.

We deliver a unified data layer and one source of truth across your systems.

Integrate · Nexus
L2

Data quality

Stuck: duplicates, gaps and no shared definitions.

We deliver diagnostics, validation and observability, so you can trust what you measure.

Diagnose
L1

Data collection

Stuck: data locked in SaaS tools, spreadsheets and manual exports.

We deliver connectors and ingestion, built on Airbyte. We own the plumbing.

Connect
The bet

Own the plumbing.

Whoever owns where the data lands and how it gets there owns every layer above it. Two primitives, everything else builds on them.

Database

One governed store that every source lands in. The single source of truth the whole business reads from, rather than ten systems each holding a version.

The foundation

Connector

Airbyte-based ingestion from your tools, systems and stores. Managed and monitored by us, running into your database on a schedule you can trust.

The pipes
How we take you up

One ladder, five steps.

Each step is a defined engagement with a clear start and end. Enter where you are. The assessment tells you where that is.

01

Diagnose

Where you sit on the pyramid, and the gaps that hold you back. Starts with the scorecard.

Start here
02

Connect

Ingestion from every source, built on Airbyte. The plumbing everything else runs on.

03

AutoML

Models trained on your own data and tuned to your metrics, without building an in-house data science function.

04

Visualise

Reporting and dashboards that stakeholders read without a translator.

05

Talk to it

Natural language over the whole estate, deployed in your environment where regulation demands it.

The short version
Connect all your data. Visualise all your data. Talk to all your data.

The three lines are levels 3 to 5 of the pyramid, in plain terms. We earn the right to each one by getting the level below it right first.

AI Readiness Scorecard

Practitioner framework. Independent diagnosis.

We publish the assessment methodology. The engagement is where the work happens. Score your gaps, then book a diagnostic when you are ready to close them.

Answer five questions, one per category. You get a maturity score out of 100, your level on the pyramid, and a prioritised list of what to fix first. It runs entirely in your browser, so nothing is sent anywhere unless you ask for your breakdown by email.

The questions and the scoring logic are public, so a firm that scores itself at Level 2 already knows what a Level 4 looks like before it ever picks up the phone. We publish the diagnostic. We deliver the build.

Take the assessment Book the paid diagnostic
  • Published methodology. Questions, scoring rubric and gap logic, available to run and to embed.
  • Score yourself. A maturity level and per-category breakdown in minutes.
  • See the gaps. Your weakest categories named plainly, with the honest next step.
  • Go deeper. Turn the gap into a costed plan with the paid diagnostic.
At enterprise scale

One score hides the unit that holds you back.

For companies with £500m–1.5bn revenue, the scorecard is the starting point for the diagnostic, not the whole answer. Three things change at that size.

Six dimensions, not five. The diagnostic adds governance and commercial alignment, because large firms have many system owners and a board that asks for evidence.

Each business unit is scored separately. Divisions sit at different levels, and a group average hides the weakest one. We report the spread and name the unit that sets the pace for everyone else.

Every answer is checked. We interview the CFO, COO and CDO, and test each answer against your systems, your board pack and sector benchmarks.

One way to reach Level 5: route AI work by evidence

  • Shortlist. Independent public leaderboards narrow each AI task to two or three candidate models on quality, price and speed.
  • Test on your own work. Each candidate runs on 20–50 of your real examples with known good answers. Public scores give a shortlist. Your data picks the winner.
  • Route each request. A fast classifier labels every request by task, difficulty and data sensitivity. A routing table built from the tests picks the model.
  • Measure and retest. Every outcome is logged and the tests rerun monthly, so the board sees what AI costs and what it delivers.

This needs Level 3 first: without trusted answers there is nothing to test against.

Find out where you stand.

Around three minutes to a maturity score you can take to your board. Book a diagnostic when you are ready to go deeper.

Take the assessment Get in touch