Seeking Guidance on AI-Assisted SharePoint Migration

Question

Seeking Guidance on AI-Assisted SharePoint Migration

BALU E S 0

I am currently working on a SharePoint migration scenario involving 11 million documents across SharePoint 2010 and SharePoint 2013, with the target platform being SharePoint Online.

The business does not want to follow a simple lift-and-shift approach. One of the key challenges is determining which content is still relevant, which documents are duplicates or near-duplicates, and which content is subject to retention or regulatory requirements.

For example:

2019 Loan Agreement

Last Modified: 2019

Modified more than 2 years ago

Document Type: Loan Agreement

Retention Category: Financial / Regulatory Record

A simple age-based migration rule might exclude this document. However, based on its classification and retention requirements, the decision could be:

MIGRATE — RETENTION EXCEPTION

This led me to explore whether an AI-assisted content reconciliation layer could be introduced during the migration rather than performing classification and reconciliation only after the content reaches SharePoint Online.

My initial thought was to explore Microsoft Syntex / Microsoft 365 content understanding capabilities to help classify documents, identify potential duplicates and near-duplicates, and surface retention-relevant content before migration, with a human-in-the-loop review for sensitive business and Legal decisions.

However, I would like to validate the architecture with the Microsoft team:

Is Microsoft Syntex / Microsoft 365 content understanding the right approach for this scenario, particularly when the source content is in SharePoint 2010/2013?

Or would you recommend another Microsoft service, architecture pattern, or combination of capabilities for:

Content classification before migration

Duplicate / near-duplicate detection

Retention and compliance identification

AI-assisted migration decisions

Human validation of exceptions

I would appreciate any guidance on whether this is the right direction or if there is a more appropriate Microsoft-native approach.

The objective is to use AI as a meaningful migration control, rather than simply adding AI to the solution.

0 comments

2 answers

Your answer

Answer 1

Pavle Peković 6,845 Volunteer Moderator

Syntex is the wrong tool for this. It is a library-bound service: models and autofill columns run against files that are already in a SharePoint Online document library.

The second reason is cost. Syntex is billed pay-as-you-go. At 11 million documents it will be more than significant cost.

You can't migrate 11M documents using the same approach for all of them. You can create rules for some of them, you can apply some rules to groups of documents. You don't need AI assistance for every single one of them.

What you need is clear staged plan. I prefer a pilot. Learn and make corrections. Go with a second wave.

Maybe it's a good idea to hire migration expert(s) and/or use 3rd party migration tool.

BALU E S 0 Reputation points

2026-08-26T13:01:51.01+00:00

Thank you, this is helpful. My initial thought was to explore AI-assisted classification because the challenge is not simply moving the content but making migration decisions based on business context and retention requirements. I agree that applying AI across all 11M documents would neither be necessary nor cost-effective. A staged approach using inventory, deterministic rules, duplicate analysis and targeted AI for exceptions, followed by a pilot and iterative waves, seems more appropriate. I was particularly interested in understanding whether there is a Microsoft-native capability for the targeted classification stage when the source content is still on SharePoint 2010/2013.
Pavle Peković 6,845 Reputation points Volunteer Moderator

2026-08-29T09:34:21.7466667+00:00

I don't think you will find one major Microsoft tool that will help you with everything. 11M documents are a lot. I would hire a migration specialist.

Answer 2

Hi @BALU E S,

Thank you for posting your question in the Microsoft Q&A forum.

Based on your description, this is a large scale migration scenario involving approximately 11 million documents across SharePoint 2010 and SharePoint 2013, with additional requirements around AI-assisted classification, duplicate detection, retention analysis, compliance review, and migration decision-making. Because of the scale and complexity involved, as well as the need to validate the most appropriate Microsoft-native architecture.

I would strongly recommend engaging Microsoft Support and the relevant migration specialists to obtain official guidance tailored to your environment. Microsoft's support and engineering teams are best positioned to provide authoritative recommendations and identify any platform limitations, best practices, or supported migration patterns.

To obtain the most accurate guidance, the Global Administrator should create a service request with Microsoft Support. For detailed instructions, please refer to Get support - Microsoft 365 admin.

As community contributors, I do not have access to internal engineering guidance, migration assessment tools, backend diagnostics, or architecture review resources. For enterprise-scale projects involving millions of documents and compliance-sensitive migration decisions, Microsoft Support is the appropriate channel to engage the relevant product teams and specialists.

I apologize for redirecting you to Microsoft Support. However, due to the size of the migration and the architectural guidance being requested, the support team will be able to provide more accurate and comprehensive recommendations than what can be offered through the community forum.

Thank you again for your time and understanding. While my initial response may not directly answer the architectural question, I wanted to ensure you are directed to the appropriate team that can provide authoritative guidance for a migration of this scale.

Note: Please follow the steps in the forum documentation to enable e-mail notifications if you want to receive the related email notification for this thread.