Data & AI Engineering · MSA Professional Services

Teaching an ERP to
document itself.

Analysts were spending hours reverse-engineering cryptic column names and chasing subject-matter experts. We automated the entire documentation pipeline — and made updating it as easy as a chat.

Microsoft Fabric Copilot Studio Power Automate SharePoint MS Dynamics 365 T-SQL
~80%
Reduction in metadata
compilation effort
2
Automated flows keeping
data in sync 24/7
1-click
Agent access from Teams,
Outlook, or any MS tool
Why trust it

Limits stated up front: tested on synthesized ERP data, performance at production scale not yet validated, and human review treated as essential on every AI-generated definition.

The problem

Data everyone has. Context no one can find.

MSA's Dynamics 365 ERP had evolved faster than its documentation. Columns had been renamed and repurposed with no record of what they meant, business logic lived in a few people's heads, and the manual data dictionary fell behind every schema change. Analysts burned hours on trial-and-error queries — and updating the docs always lost out to billable work.


How it works

Five tools. One seamless pipeline.

Everything runs inside MSA's existing Microsoft 365 stack — no new vendors, no new training. Fabric pipelines extract SQL metadata from Dynamics 365 on a daily schedule and upsert it into a SharePoint data dictionary, while a Copilot Studio agent lets anyone look up or update business definitions in plain language from Teams or Outlook — with Power Automate writing validated changes back in real time. And every AI-generated definition passes through human review before it's saved: the agent drafts, a person decides.

End-to-end metadata pipeline
🗄️
Dynamics 365
ERP source
🏗️
Fabric Lakehouse
Raw storage
⚙️
Fabric Warehouse
SQL + stored procs
📋
SharePoint
Data dictionary
🤖
Copilot Agent
Natural language

Sample output

Documentation that feels like a conversation.

A real interaction from the prototype.

Metadata Documentation Agent · M365 Copilot
👤
Actual amount
🤖
The closest match is ActualAmount.
Current definition: "Represents the actual amount recorded for the transaction."

Would you like to update this definition?

Reflection
Behind the build

The hard part wasn't the pipeline — it was getting busy professionals to contribute to documentation without it feeling like a burden, and the answer was to make it a conversation inside the tools they already use. The prototype ran on synthesized ERP data, and the occasionally generic AI definitions confirmed that human review stays essential. The best AI tools don't replace human judgment. They make it easier to express, capture, and preserve it.

Interested in working together?

Get in touch