Session: How SharePoint can help you maximize the value of your content processing
Speaker: Chris Bortlik, Kristen Kamath, and Sanjoyan Mustafi
Content is the lifeblood of every organization but managing it effectively can be challenging. Traditional methods of content management are often manual, error-prone, and inefficient. That's why SharePoint offers a set of cloud content management capabilities that leverage AI to help you organize, process, and govern your content with ease and speed. “Content is what makes copilot what it is” – Chris Bortlk
I was lucky enough to attend the Microsoft 365 Collaboration Conference 2024 session by Chris Bortlik, Kristen Kamath, and Sanjoyan Mustafi, who demonstrated how SharePoint and its new premium features can help you maximize the value of your content with AI-powered content processing. The approach for the session was not your generic feature list but utilized a collection popular business scenarios that can be improved using content AI, Premium, and more. They walked through each scenario and demonstrated how what an optimized solution could be then walked backwards for what solutions were put together to make it work.
What is AI-powered content processing?
AI-powered content processing is a set of features and tools that SharePoint provides to help you analyze, extract, and act on the information in your documents. It uses advanced AI models to understand the content and context of your files, and automatically apply metadata, tags, translation, OCR, and more. It also enables you to streamline your content-centric workflows, such as collecting signatures, generating documents, and annotating files.
They had a good driving point during the session about the size and complexity of unstructured data in M365 that drove home the importance of solutions they went through. By 2025, 80-90% of data growth will be associated with unstructured data. With this incredibly large amount of unstructured data, which Chris stated now includes the addition of ~2.5 billion documents a day into M365, leads to challenges in discoverability, security, and retention to the content. They then began to break down where AI can help.
Existing challenges with AI-powered content processing
In this session they introduced 3 key challenges or scenarios that could be improved with content AI which included:
- Information overload
- Chris talked about how this is the challenge of “why you should tag it” without an incentive. You can use AI to discover and reuse knowledge from your documents, without having to read through and classify large sets of documents manually.
- Lost information
- You can use AI to connect and manage your content, and improve its security and governance with integration to Microsoft Information Protection. Chris talked about how even not being able to find content can be considered lost information!
- Loss in productivity
- You can use AI to automate the capture, ingestion, and categorization of your content, and reduce the time and effort spent on manual data entry and tagging. This was all about how manual data entry delays actual processing.
What are some of the scenarios where AI-powered content processing can help?
Next, they showcased three scenarios that Kristen started and went through where AI-powered content processing can help you optimize your content management and business processes. This included:
- Business process automation
- This was my favorite demo as I see it as the top scenario I am seeing in the real world. Kristen went through how you can use AI to process invoices, receipts, contracts, and other common documents, and extract key metadata pairs, such as total amount, due date, vendor name, etc. She also showed how you can use setup autofill columns, trigger workflows, annotate files, assemble documents, merge and extract PDFs, and query content based on metadata. This was full of demos about ways to extract and utilize metadata so you can turn your library experience into something like this!
- Business process automation
- Kristen highlighted prebuilt models that exist including invoices, receipts, contracts to get started in this space while showing the coolest part of the demo for new autofill columns which actually uses LLM logic to fill metadata!
- She ended with showing how OCR is utilized to pull info from an image-based document including and that content query is what is actually used to search for information in the library easier such as finding invoices over $5000.
- Content migration
- Next up was content migration which focused on how you can use AI to organize your content when you migrate it from other sources, such as file shares, legacy systems, or third-party cloud services. You can also use AI to classify, summarize, identify the language, and apply taxonomy tags to your content when it gets to M365. Additionally, show that you can use AI to translate your content into over 100 languages, and OCR to make your image-based files searchable and governable. Kristen explained how you should make your migrations “smarter” by preconfiguring the destination document libraries with something like this:
- Content migration
- Kristen showed how you can use taxonomy tagging which can automatically apply tags from a companies managed term set and how to use OCR for post migration for DLP policies to reduce oversharing.
- She then showed how you can translate documents with over 100 different languages directly in SharePoint and then trigger built in rules, similar to the drop off folder experience we’ve had in the past!
- Collecting signatures
- Sanjoyan took the stage next to talk about how you can use the new SharePoint eSignature solution to request and sign documents directly within the Microsoft 365 ecosystem, without having to use extra logins or applications. You can also integrate with Adobe Sign and DocuSign, and leverage SharePoint's advanced content management tools, such as content query, content assembly, and Microsoft 365 Archive to build solutions around colleting signatures.
- SharePoint eSignature is a first party provider within M365 and the beautiful part is it is starting from within a document library in SharePoint and the document ends up back in the same folder after signature, which includes permissions just like where you started.
- Sanjoyan showed how you could imitate this process through Teams as well while showing how you can see the document stages of where the document is at in its signing lifecycle.
- He ended with details on pricing and roadmap which they key points were if you don’t use it, you don’t pay for eSignature! This is in the process of deployment across geos, commercial, and GCC tenants in which there are nuances of who has what, but the plan is to have it globally available in 2025. Some of the nuances are highlighted in the eSignature roadmap below.
What is the roadmap for AI-powered content processing?
The session also shared a small roadmap of the coming soon features in this space highlight the ones I am most excited for with Autofill columns and Sensitive information detection.
This was a great session that highlighted real-world scenarios for using content AI and didn’t feel like new feature blitz. There are so many tools and understanding how they work together is key in figuring out what you can do within your own organization. Great job Sanjoyan, Kristen, & Chris!
