Consulting Website + Client Portal Development (Fabric)

Thanks for your question, our vision is:

1️⃣ Client Login

Clients visit portal.datanovalabs.com (hosted in UAE).

They log in securely via Microsoft Entra ID (Azure AD B2C) — Fabric


2️⃣ File Upload

Inside their workspace, clients click “Upload Data”.

They can attach files such as .csv, .xlsx, SQL bulk, .json, zip etc

Each client has their own secure storage container in the Fabric environment — completely isolated from other clients.

Clients can upload multiple pieces of information and can create new folders, as and when required, however there will be a folder called "Reporting Raw Data" where all files within that folder will be cleansed and then reported upon using Power BI. The other folders/ files which clients create/ upload are used only for hosting, not reporting, we will not do anything with them.

OneLake acts as the central data lake for all uploads.

3️⃣ Automated ETL Trigger

Once uploaded:

1. A Fabric Data Pipeline detects the new file.

2. It launches a Fabric Dataflow Gen2 or Fabric Notebook (Python) to:

Clean data (remove blanks, duplicates, errors).
Standardise field names and formats.
Merge or join data where needed.


3. The cleansed dataset is saved back to OneLake as:

/OneLake/DataNova/Clients/{ClientName}/Processed/{DatasetName}.parquet



4️⃣ Reporting Layer

The processed data feeds into Power BI semantic models within Fabric. Developed by external.

Each client has a workspace or dataset tagged with their tenant ID.

Reports auto-refresh whenever a new file is uploaded and processed.

Clients can view dashboards directly inside the portal’s embedded Power BI viewer (via Power BI Embedded).


5️⃣ Notifications & Audit Trail

Power Automate or alternative sends a confirmation email or Teams message to:

The client (“Your data upload has been successfully processed”).

Your internal team (“New upload from Client X at [timestamp]”).


Audit log captures:

Who uploaded

When processed

What scripts were executed

Output file names



All this is stored in Fabric’s activity log and optionally mirrored in Azure Log Analytics.


6️⃣ Optional Enhancements

✅ Use Fabric Copilot for automated anomaly detection or predictive summaries inside reports.

Hi Pradip,

Thanks for sharing such a clear end-to-end workflow. Your Fabric setup with Entra ID, OneLake storage isolation and automated ETL is well-defined and definitely achievable. To move ahead, could you share whether you also want us to handle the full portal UI/UX design, or will you provide the frontend screens on your side?