Inerra

Architectures

Patterns we have designed and delivered.

A selection of the architectures we have delivered, drawn with the components that are actually in them. Filter by what you are trying to build, and open any one for the full case.

Data platform

Medallion lakehouse on Microsoft Fabric

A charge point operator growing fast, with a Power BI import model breaking under the weight of it: large files, slow refreshes, and one person who understood how it held together. We replaced it with a full medallion lakehouse on Fabric and owned every architectural and cost decision. It later absorbed a full CPMS migration with no structural rework, because the separation of concerns was right first time.

  • Microsoft Fabric
  • OneLake
  • Delta Lake
  • PySpark
  • Azure Functions
  • Power BI

Sources

CPMS
OCPP messagesCharger telemetry
Billing and CRM

Ingest

Fabric Data PipelinesBatch sources
Azure FunctionsReal-time OCPP

Bronze

OneLake / DeltaRaw, append only

Silver

PySpark notebooksDedup, nested struct coercion
Delta MERGE upserts

Gold and use

Star schema
Power BI
Fault alerts to engineers

Delivered: real-time OCPP ingestion for proactive charger fault detection, cutting engineer response time to critical errors. Client named on request.

Read the case study →
Data platform

Secure enclave inside an enterprise data mesh

The sensitive and secret data domain of a global energy supermajor's data mesh, holding PII, sensitive and classified data at organisational scale. SOX and GDPR were designed into the architecture rather than bolted on afterwards, and the controls sit at the boundary, so a team that needs the data does not have to stand up a platform of its own to hold it.

  • Azure
  • Data mesh
  • Data tokenisation
  • Custom RBAC
  • Private Link
  • Key management
  • Azure DevOps

Restricted sources

Sensitive domain data
Classified data
PII

Boundary

Private LinkNo public path in
Data tokenisation
Access approval

Enclave

Isolated storageCustomer-managed keys
Governed compute
Custom RBAC

Published

Governed data product
Full audit trail

Consumers

Approved teams only
SOX and GDPR in the design · risk register kept to audit standard · roadmap owned across multiple quarters

Delivered: led the security engineering team on the design and build. Every stage approved by a Security Architectural Review Board of more than ten members at VP and SVP level. Client named on request.

Read the case study →
Data platform

Governed ingestion on a UK insurance data platform

Taken over as architect mid-engagement. Every new source had been arriving as its own bespoke pipeline, so we standardised the ingestion pattern and put a proper governance layer on Unity Catalog underneath it. Onboarding a new data source took half the time afterwards.

  • Databricks
  • Unity Catalog
  • Delta Lake
  • Medallion
  • Azure DevOps

Sources

Line of business systems
Third party feeds

Ingest

Standard ingestion patternOne shape, every source

Lakehouse

Bronze
Silver
GoldDelta Lake on Databricks

Governance

Unity CatalogCatalogue, lineage, access

Consumers

Analysts
Models

Delivered: new data source onboarding time halved. Client named on request.

Read the case study →
IoT architecture

Distributed generation platform, one site to 50 MW

A multi-tenant platform for renewable and industrial generation across some of the UK's largest manufacturers, built so a new site or a new generation technology could be onboarded without major development. The commercial model carried no CAPEX, so cloud cost per site had to keep falling as the portfolio grew.

  • Azure IoT Hub
  • IoT Edge
  • Stream Analytics
  • Azure Data Explorer
  • Azure Digital Twins
  • SignalR

Plant

CHP engines
Solar inverters
PLCsPlant control
MetersModbus TCP/RTU

Edge

Edge modulesGeneration against site demand, export margin, grid events

Ingest

Azure IoT Hub

Process and store

Stream Analytics
Azure Digital TwinsSite and asset model
Azure Data ExplorerTime series telemetry

Use

SignalR dashboardsReal-time
Grid balancing

Delivered: grew from a single CHP site to over 50 MW of live generation, with cloud cost per site trending down as the portfolio grew.

Read the case study →
IoT architecture

Water quality monitoring across field sites

Distributed field sites for a supplier to the national water utilities. Water quality sensors on remote hardware, their readings brought back over Azure IoT Central, processed and turned into something the business could report on. The hard part was reliable data off scattered sites, not the stack.

  • Azure IoT Central
  • Azure Functions
  • Stream Analytics
  • Azure Data Lake
  • Power BI

Field

Water quality sensors
Embedded firmwareC

Connect

Azure IoT CentralDevice connectivity

Process

Azure Functions
Stream Analytics

Store

Azure Data Lake

Use

Power BI
Read the case study →
Edge & integration

Asset tracking for equipment hire

GPS on the plant, BLE tags on the small kit that goes missing, and a booking lifecycle that knows what is out, where it is and what it is earning. Multi-tenant from the first commit, for the UK and the UAE. We run this one ourselves, as Equipy.

  • Azure Container Apps
  • Azure SQL
  • Row level security
  • Entra ID
  • Azure Maps

Assets

GPS trackers
BLE tagsTracker acts as the gateway

Ingest

Container Apps
Device registry

Store

Azure SQLMulti-tenant, row-level security
Position history

Application

Booking lifecycleAudit trail on every state change
Entra ID auth

Use

Azure Maps live view
Utilisation and revenue
Read the case study →
Applied AI

AI-assisted property inspection

The architecture behind Tamaneena, our own product. An inspector walks a unit and photographs the defects, and the platform turns the images, the room and defect context and the notes into a contractor-ready snagging report. A person signs off, and every step leaves an audit history. Practical AI on a governed platform, not a demo.

  • Image analysis
  • Defect classification
  • Description generation
  • Human review
  • Audit history

Inputs

Site photographs
Room and defect context
Inspection notes

AI processing

Image analysis
Defect classification
Description generation
Severity and trade suggestion

Application

Human reviewA person signs off
Contractor-ready report
Audit history

Delivered: live as Tamaneena in the UAE, proof that we build AI applications that reach production.