Services

Six disciplines, one chain.

From the field record to the decision: databases, geographic information systems, analysis, models, and the people who run them.

01 — Fields of activity

Six disciplines, one chain.

A model is only as good as the analysis behind it, the analysis only as good as the database beneath it, and the database only as good as the people who run it. We work across the whole chain rather than one link of it.

Design and administration

Database design & management

From the data model to daily operation. We design the structure first, enforce the rules at the point of entry, and run the system afterwards if you would rather not staff it internally.

  • Relational and spatial data models
  • Enterprise geodatabases in PostgreSQL/PostGIS and SQL Server
  • Migration and consolidation of legacy and historical datasets
  • Validation rules, integrity constraints and QA/QC reporting
  • Scheduled loading, backup and administration
  • Exploration and geological databases, including DataShed
Architecture and deployment

GIS design

Spatial data infrastructures built to be handed over: documented, standards-based, and operable by your own staff after we leave.

  • Spatial data infrastructure design and governance
  • WebGIS platforms and mapping applications
  • Satellite imagery processing, optical and radar (SAR)
  • Terrain, hydrological and watershed modelling
  • Cartographic production and map series
  • Metadata, catalogues and OGC-compliant services
Method and interpretation

Data analysis

The step between holding a dataset and knowing something. Analysis is documented and reproducible, so a result can be re-run and defended rather than taken on trust.

  • Exploratory and descriptive statistical analysis
  • Spatial statistics, interpolation and geostatistics
  • Time series, trend and anomaly detection
  • Change detection from satellite imagery
  • Indicator frameworks and periodic reporting
  • Reproducible analysis in Python, with the code handed over
Engineering at scale

Big data

Where the volume, the variety or the rate of change puts a problem beyond a desktop tool. We build the pipeline that turns raw input into something a decision can rest on.

  • Automated ingestion and processing pipelines
  • Large raster and vector processing at scale
  • Machine learning and deep learning on spatial data
  • Model training, validation and deployment
  • Dashboards and decision-support interfaces
  • Monitoring, early warning and alerting systems
Subsurface and resources

Geological modelling

Drilling data turned into a defensible picture of what is in the ground — and an estimate that will survive review by a competent person.

  • Implicit and explicit 3D geological modelling
  • Wireframes, solids and cross-sections
  • Variography and geostatistical estimation
  • Block models and grade interpolation
  • Resource classification aligned with JORC and NI 43-101
  • Leapfrog, Micromine, Surpac, Datamine and Geosoft
Capacity and continuity

Training & assistance

Two things usually treated as afterthoughts and should not be: teaching the team to use the system, and staying reachable once it is in production.

  • Intensive courses in French, Arabic and English
  • Curricula built around your own data, not sample datasets
  • On-site or remote delivery, certificate issued
  • Technical assistance contracts with agreed response times
  • System documentation and handover packs
  • Ongoing support for systems we did not build

OUR EXPERTISE, ILLUSTRATED

Make complex information readable.

Combine spatial layers

ObservationsGeologyTerrain
Overlay field observations, geological units and terrain to interpret spatial relationships.

Understand the subsurface

DH01DH02DH03Boreholes + geological interpretation
Connect borehole observations with interpreted geological units to build a consistent model.

Conceptual diagrams; no project data. Photographic illustrations generated with AI.

02 — Types of engagement

Three forms of engagement.

Every assignment begins with a written scope: what is to be delivered, in what form, and against which criteria it will be accepted. The form of the engagement follows from the nature of the work, not the other way round.

Continuous operation

Operation & support

For a database or a platform that has to stay available. We take on its day-to-day operation alongside your own team, under an agreement that sets response times and reporting obligations on both sides.

  • Agreed response times for loads, imports and incidents
  • Quality reporting issued on a fixed cycle
  • Periodic review of the arrangement with your team
  • Escalation path and named points of contact
Defined deliverable

Deliverable-based assignment

For a specified result: a geodatabase, a WebGIS platform, a processing pipeline, a satellite analysis, a set of map products. The engagement is defined by the deliverable and by its handover.

  • Written scope, milestones and acceptance criteria
  • Source data, code and documentation handed over
  • Built so your own staff can operate it afterwards
  • A period of support after handover
Specific expertise

Training & expert assignments

For training sessions, and for assignments carried out against terms of reference — including as the registered Mauritanian partner within an international consortium.

  • Courses delivered in your offices or ours
  • Assignments framed by terms of reference
  • Available for tender submissions as prime or partner
  • Second opinion and technical review of existing work

04 — Sectors

Where this work is needed.

The discipline is the same across all of them. What changes is the vocabulary, the regulatory framework and the consequence of getting it wrong.

Mining & explorationdrilling · assay · resources
Water & hydrologynetworks · basins · resources
Agriculture & landland use · degradation
Environmentmonitoring · impact
Public administrationcadastre · planning
Risk & food securityearly warning · indicators

05 — Stack

Tools in daily use.

We work on your licences where your workflow requires it, and on open-source tools where it does not — and pick the tool that fits the task rather than the one already installed.

PostgreSQL / PostGISSQL ServerDataShed ArcGIS ProQGISGeoServer ENVIERDAS ImagineSNAP Sentinel-1 SARGlobal MapperMicromine LeapfrogSurpacGeosoft PythonTensorFlowPower BI

06 — Questions we get asked

The things people check before they sign.

Mostly asked by managers who have never placed a database or a platform outside their own organisation.

How does remote database management actually work?

Your field teams, laboratories and departments send data the way they already do. We load it, validate it against the rules agreed for the project, flag what fails, and return it to the named owner for correction. You receive quality reporting on an agreed cycle and exports in the formats your own software expects.

In practice the person entering the next record never has to think about it. That is the point of the service.

Who owns the data, and how is it kept secure?

You do. We work inside your systems wherever possible. Where a local working copy is unavoidable, it is encrypted, backed up on a documented routine, and destroyed at the end of the engagement on your instruction.

A non-disclosure agreement is signed before any data moves. For listed companies, exploration and production results are price-sensitive and are treated accordingly.

What standards do you work to?

Metadata follows ISO 19115. Services are published to OGC specifications so that any compliant client can consume them. Where a deliverable feeds mineral resource reporting, the data governance is aligned with the requirements of JORC and NI 43-101.

Where a client has its own internal standard, that takes precedence and we work to it.

Whose software licences are used?

Yours, in most operational engagements — it is the normal arrangement and avoids a duplicate installation. We hold our own licences where a deliverable requires it, and default to open-source tools such as QGIS, PostgreSQL/PostGIS, SNAP and Python wherever they do the job as well.

Do you work on site or remotely?

Both. Database operation, platform development and data processing are delivered remotely, which is what makes continuous availability practical. Training, system commissioning and field data collection are delivered on site.

Nouakchott is on GMT year-round, which puts the working day in overlap with West Africa and Europe for its full length, with the Gulf until mid-afternoon, and with Western Australia early in the morning.

Which languages do you work in?

Arabic, French and English. Reporting, training material and documentation are delivered in whichever of the three the client works in.

Can you act as the local partner on a Mauritanian tender?

Yes. Richat Data Science is a Mauritanian company with resident senior expertise, which addresses the local-content requirement in national procurement and removes the mobilisation an international bidder has to organise. We are available as the named local partner in a consortium, or as prime contractor.

Send the shape of the problem.

Start a conversation