Problems solved, told as problem → solution → outcome

The problem

Structured files (CSV and Excel) needed to land in Dataverse reliably, at variable volumes, without a developer touching code for every new file format.

Our approach

Blob Storage triggers publish file metadata to Service Bus; an Azure Function acts as a generic import engine driven by a declarative configuration (a small DSL). Retries and dead-lettering protect against data loss, and file routing archives or quarantines each file by outcome.

The problem

Manual solution exports and imports were slow, error-prone and impossible to audit across dev, test and production.

Our approach

Source-controlled solutions with Azure DevOps build and release pipelines: automatic solution version incrementing, packaging as managed solutions, and gated deployment to each target environment — including a custom task to merge multiple solutions into one release artifact.

The problem

Out-of-the-box controls couldn't express the interaction a team needed, slowing down everyday work.

Our approach

React and Fluent UI PCF controls bound directly to Dataverse, styled to match the platform, with business rules enforced server-side by plugins — packaged and deployed through the same solution pipeline.

The problem

Heavy processing and third-party integrations didn't belong inside the platform and needed to scale independently.

Our approach

Azure Functions expose custom APIs and process webhooks and events, decoupled from the platform via Service Bus, with observability through Azure Monitor so failures surface loudly.

The problem

Business rules that must always hold couldn't live in the client — they needed to be enforced on the server.

Our approach

Well-structured plugins and custom APIs run inside the platform pipeline, validating data, automating tasks and raising events for downstream integration — all delivered under proper ALM.