Event banner
Event details
97 Comments
- By having Sharepoint 365 and Sintex installed in the same domain or environment, will SP inherit all the features of SYNTEX?
- IanStory
Microsoft
Hi Jaime, each user of Microsoft 365 that is using Microsoft Syntex needs to have a Microsoft Syntex license if they want to use features like Content Query, Content Assembly, etc. unless the only features they want are unstructured document processing and/or prebuilt document processing, which are also about to be available as pay-as-you-go starting on March 7 (as noted on the blog - https://techcommunity.microsoft.com/t5/microsoft-syntex/microsoft-syntex-pay-as-you-go-will-be-generally-available-on/ba-p/3749868) - once these features are enabled for pay-as-you-go, any user of Microsoft 365 can use them, and their usage will be charged per use, instead of requiring a per user Microsoft Syntex license for those two features.
Is there currently a way to programmatically interact with stored annotations? If not, is there a way planned?
- IanStoryNot currently, but we plan to build an API for this, see the question and response on the other question you asked 🙂
- Alright, thanks for your answer.
- Is there any documentation on how the accuracy/confidence score is calculated? Is there a way to validate its operation?
- JamesEcclesI assume you are referring specifically to unstructured document processing. The accuracy score during model training is based on the number of predictions that match the labels in the training set. The relevance of this score will therefore be driven by how many files you have in the training set and how diverse they are. The confidence score shown in a library is broadly a product of the proximity that a file has to the all of the labels and explanations in a classifier. A tip here is to not treat this as an objective score. In real world scenarios the score may or may not give an accurate reflection of whether the outcome of the model is successful. We hide the confidence score by default, because it can be misleading as to the success or failure of the model.
- As a partner, how can we help Syntex improve its performance?
- JamesEcclesIf you mean with model accuracy, then I would say the best thing a partner can do is to empower customers to be self service on model building. The key knowledge when training a model is an understanding of the content, and the conventions in that content. The more that domain subject matter experts are involved in the process, the better the model quality.
- Thanks for your James! it is always a pleasure!
- What are the main business use cases for which Syntex brings a strong added value and which should be promoted to our customers to make them want to adopt Syntex?
- IanStoryHi Christophe! I'd add to the excellent reply from James that Syntex is taking the manual process of classifying documents, extracting metadata from documents, applying retention and sensitivity labels to documents, etc. and removing the majority of the manual work. In turn, this makes it much, MUCH easier to find the documents you're looking for (and we even provide a more powerful search experience, with the Content Query feature of Syntex), be compliant with things like records management rules and regulations, and automate business processes using the metadata extracted. Beyond this, Syntex provides end user features like universal annotation that make it much easier to collaborate (on PDF, TIFF and more files to come soon) and generate new documents (using content assembly). In all, Syntex is best at helping customers handle higher volume content processes, or manage large estates of content that are currently unmanaged (or somewhat unmanaged). We've of course also announced a broad range of new features coming to Syntex at Ignite last year, from backup and archive to eSignature, summarization, and translation - each of which has many benefits/use cases 🙂
- JamesEcclesWe have an article that describes some common use cases here- https://learn.microsoft.com/en-gb/microsoft-365/syntex/adoption-scenarios?view=o365-worldwide. This article is in the process of being updated, so keep an eye for a refresh in the coming weeks. You could also take a look at our public case studies for Syntex - https://customers.microsoft.com/en-us/search?sq=%22Microsoft%20Syntex%22&ff=&p=0&so=story_publish_date%20desc
- Do you have recognition of barcodes and QR?
- IanStoryNot today, this is on our backlog - AI Builder and Forms Recognizer have this, but we have to get it enabled on Syntex - we do intend to do this work, but I don't have a timeline to share at the moment.
- If ChatGPT is integrated, could it be used to classify documents based on their content or structure? For example, it can be trained to distinguish between different types of invoices, purchase orders, medical documents, etc. Or design graphically and without training a model, the keywords and values that you want to extract, either by relationship or proximity. Is it contemplated?
- IanStoryHi Jaime - we don't plan to integrate ChatGPT into Syntex - ChatGPT is a product made by OpenAI and there are ramifications (such as sharing data in your tenant with OpenAI's product) to integrating it directly. However, the GPT technology behind ChatGPT is something that Microsoft and OpenAI are partnered on, as you've surely seen in the news, and we continue to look at what we might be able to do with Syntex here. Stay tuned on timelines, there's nothing I can share at the moment there. In terms of the overall capability of GPT (and yes, ChatGPT) it can distinguish between different types of invoices, purchase orders, medical documents, etc. and it can also extract keywords and values. We're looking at the best way to incorporate this tech 🙂
- Can syntex be used to maintain a government-mandated classification scheme for organizations?
- JamesEcclesSyntex models can be used to add a retention or sensitivity label to processed files. These labels can be aligned to existing classification schemes. Note though, that Syntex models would be created for types of files (e.g. Contract, Invoice etc.) which you could then associate with the appropriate classification
- Thanks James!
- Hi, any update on when we can expect e-signature to become availe for testing / use?
- IanStoryHi Benoit, we are making great progress on eSignature as announced at Ignite, and we will have more details on availability soon. Our current target remains to have it around the end of the first half of the calendar year, but these things can change as we continue development and testing.
- Can syntex be used to provide a thematically grouped view of documents to users? No matter where the documents are located. The topic would be identified by the AI.
- I've done this for some clients. The approach I took was to have Syntex extract one or more "categories" of metadata from the content. For eample, you can have region values of Americas, Europe, Asia, etc. Then use search webpart or something similar to organize documents using the values extracted. Similar to metadata filter or navigation.
- Many thanks Mario!
