---
title: "Data"
updated: "2026-11-16"
url: "https://www.etheon.com/services/data"
---

# Data engineering and analytics

Data engineering, from pipelines to BI. We build the pipelines, platforms and governance behind your analytics and your AI.

order 1042

### 01 Sources

Sources.

Every order starts as a line in the export of the system that took it.

```
orders export
1042,C-88,49.00,EUR,2026-11-12 09:14
```

- Customer name masked for analysts
- Lineage raw.orders → stg\_orders → fct\_orders
- Kept 24 months, then deleted

### 02 Ingest

Ingest.

The line lands in the raw layer as it arrived, each field named.

```
raw.orders
order_id     1042
customer_id  C-88
amount       49.00
currency     EUR
ordered_at   2026-11-12 09:14
```

- Customer name masked for analysts
- Lineage raw.orders → stg\_orders → fct\_orders
- Kept 24 months, then deleted

### 03 Clean and check

Clean and check.

Types and time zones fixed, and every field checked before it moves on.

```
stg_orders
order_id     1042                  unique
customer_id  C-88                  not empty
amount       49.00 EUR             positive
ordered_at   2026-11-12 09:14 UTC  in range
```

- Customer name masked for analysts
- Lineage raw.orders → stg\_orders → fct\_orders
- Kept 24 months, then deleted

### 04 Model

Model.

Warehouses and lakehouses set up for your scale, your cost and your access rules.

- The platform that fits your stack
- Costs you can see
- Access by role

```
fct_orders
order_id     1042
customer_id  C-88
amount       49.00 EUR
ordered_at   2026-11-12 09:14 UTC
segment      SMB   from customers
region       EU    from accounts
```

- Customer name masked for analysts
- Lineage raw.orders → stg\_orders → fct\_orders
- Kept 24 months, then deleted

### 05 Serve

Serve.

The order is counted into a metric your reports use and a feature your models read.

```
metrics, features
metric   weekly_active_customers  counted
feature  orders_last_7_days       3
```

- Customer name masked for analysts
- Lineage raw.orders → stg\_orders → fct\_orders
- Kept 24 months, then deleted

### 06 Decide

Decide.

People read the metric in a report; a model reads the feature as one input of many.

```
report, model
report  Weekly active customers
model   churn risk, one input of many
```

- Customer name masked for analysts
- Lineage raw.orders → stg\_orders → fct\_orders
- Kept 24 months, then deleted

### Pipelines. Data moved from its sources, cleaned and checked on the way.

#### What it is

A data pipeline moves data from where it is created to where it is used, and checks it on the way. We connect your sources, load data on a schedule or as it happens, and clean and test each record before your reports and models read it.

#### What we do

- Sources connected.
- Data ingested on a schedule or as it happens.
- Cleaning and checks on the way.

#### How we do it

1. Connect: Your sources connected, each with a named owner.
2. Load and check: Data loaded on a schedule or as it happens, checked on the way.

#### What you get

- Pipelines in your repository
- Data checks that run on every load
- A map of every source and its owner

#### Stack and tools

- Data integration
- Workflow orchestration
- Data quality checks
- Streaming

#### Related

- [Data platforms](https://www.etheon.com/services/data?detail=platforms#platforms)

### Data platforms. Warehouses and lakehouses set up for your scale, your cost and your access rules.

#### What it is

A data platform is where your company's data is stored, modelled and queried. We set up warehouses and lakehouses that fit your stack, your scale and your access rules, with the costs visible so the platform stays affordable as it grows.

#### What we do

- The platform that fits your stack.
- Costs you can see.
- Access by role.

#### How we do it

1. Choose: The platform that fits your stack, your scale and your cost.
2. Set up: Storage, models and access rules, written as code.

#### What you get

- A warehouse or lakehouse set up as code
- Access rules by role
- A cost report

#### Stack and tools

- Data warehouses
- Lakehouses
- Infrastructure as code
- Cost monitoring

#### Related

- [DevOps and SRE](https://www.etheon.com/services/cloud?detail=devops#devops)

### ML data. The datasets your models learn from, labelled and versioned with the model.

#### What it is

ML data is the data your models learn from and are tested on. We label datasets for the model's task, redact personal data before labelling, and version each dataset with the model it trains, so any result can be traced back to its data.

#### What we do

- Datasets labelled for the model's task.
- Every dataset versioned with its model.
- Personal data redacted before labelling.

#### How we do it

1. Label: Datasets labelled for the model's task, personal data redacted first.
2. Version: Each dataset versioned with the model it trains.

#### What you get

- Labelled datasets
- Versions linked to each model
- A record of how each set was made

#### Stack and tools

- Labelling tools
- Dataset versioning
- Data quality checks
- Redaction tools

#### Related

- [MLOps](https://www.etheon.com/services/ai?detail=mlops#mlops)

### Analytics and BI. Every metric with one definition, an owner and its tests.

#### What it is

Analytics and BI turn data into numbers your teams agree on. We give every metric one definition, an owner and tests that run with every refresh, and build the reports and dashboards your teams read.

#### What we do

- One definition for every metric.
- An owner for every metric.
- Tests that run with every refresh.

#### How we do it

1. Define: One definition and one owner for every metric.
2. Test: Tests that run with every refresh, before anyone reads a number.

#### What you get

- The metric definitions, in the format shown on this page.

#### Stack and tools

- Data transformation
- BI dashboards
- Metric definitions
- Data quality checks

### Governance. Access by role, and data kept in the region you choose.

#### What it is

Data governance decides who can see which data, where it is kept and for how long. We set access by role, keep data in the region you choose, mask personal data where it is not needed, delete it on a schedule and log every access.

#### What we do

- Data kept in the region you choose.
- Access by role.
- Personal data masked where it is not needed.
- Retention and deletion on a schedule.
- Every access logged.

#### How we do it

1. Map: Each dataset, who needs it and where it must stay.
2. Enforce: Roles, masking and deletion set in the platform, every access logged.

#### Questions

**Where is our data stored?**

In the region you choose. Your data and its backups stay in that region.

**Who can access our data?**

Only the roles you approve. Access is granted by role, kept to what each role needs, and every access is logged.

#### What you get

- An access map by role
- Retention and deletion rules
- Access logs your auditors can read

#### Related

- [Your data at Etheon](https://www.etheon.com/company/trust#your-data)
- [Data processing agreement](https://www.etheon.com/legal/dpa)

## What you keep.

Every metric and every dataset comes with its definition.

### Weekly active customers.

Metric · Analytics and BI

Customers with at least one order in the week, each counted once.

- Owner: Finance
- Source: fct\_orders
- Refresh: Daily, before the working day
- Tests: Not empty, one row per week

An illustration of the format, not a client's figures.

### Support tickets, labelled.

Dataset · ML data

The set an assistant is trained and evaluated on, versioned with the model.

- Labels: Intent, one label per ticket
- Version: v7 · with model v7
- Freshness: Weekly
- Personal data: Redacted before labelling

An illustration of the format, not a client's data.

## Who sees what.

Access by role, and data kept in the region you choose.

EU: Data and backups stay in EU regions.

US: Data and backups stay in US regions.

| Role | Customer identity | Orders | Support tickets | Events | Aggregates |
| --- | --- | --- | --- | --- | --- |
| Analyst | Masked | Full | Masked | Full | Full |
| Data engineer | Masked | Full | Full | Full | Full |
| Support agent | Masked | Full | Full | None | None |
| ML model | None | Full | Masked | Full | None |
| Executive dashboard | None | None | None | None | Full |

Analysts. See orders, events and aggregates in full, with customer identity and ticket text masked.

Data engineers. See every dataset to build and fix the pipelines, with customer identity masked.

Support agents. See orders and tickets to answer a customer; names and emails are masked, and every lookup is logged.

ML models. Read orders and events to learn from, with ticket text masked and no customer identity.

Executive dashboards. Show aggregates only: totals and trends, with no single customer's record.

- [AI](https://www.etheon.com/services/ai): Agents, models and MLOps
- [Cloud](https://www.etheon.com/services/cloud): Cloud, DevOps and SRE