
Microsoft 365 atWork; Senior Digital Advisor at Predica Group
Szymon Bochniak (365 atWork) published a practical video that walks viewers through using OCR and mobile scanning inside Microsoft 365 and the Microsoft 365 Copilot mobile app. In the clip, he shows how a smartphone transforms into a document capture device that feeds image-based text into Microsoft’s search and AI tools. Consequently, users who work with contracts, invoices, or handwritten notes can make those records readable and actionable for Copilot and other services.
The video emphasizes that this OCR capability is available for PDF files stored in SharePoint and OneDrive as well as for paper documents scanned on the go. Moreover, Bochniak demonstrates how to configure OCR to add text metadata so that search and Copilot recognition improve. Therefore, the feature converts previously invisible content into searchable assets without extensive rework of an organization’s information architecture.
At a technical level, OCR extracts characters from image-based documents and then indexes that extracted text so Microsoft Search and Copilot can use it. Furthermore, Microsoft manages OCR as part of its document processing services and offers it under a pay-as-you-go model that administrators can enable or limit by site. As Bochniak shows, admins can control where OCR runs, which helps align scanning with governance and storage policies.
In addition, the video highlights that recent updates make OCR more tightly integrated with Copilot workflows, including support for scanned PDFs in Copilot Chat. Thus, Copilot becomes better at analyzing common enterprise file formats that previously contained locked text. As a result, OCR now plays a role in a broader intelligent document strategy rather than acting as a standalone tool.
First and foremost, enabling OCR improves data discoverability by making scanned content searchable across Microsoft 365. Consequently, Copilot can answer questions using text that was once trapped inside images or scanned PDFs, and users spend less time on manual transcription. Furthermore, indexed text improves search relevance and helps compliance teams locate evidence more quickly during audits or investigations.
Second, the capability supports many real-world formats such as receipts, forms, contracts, and archived documents, which means organizations unlock immediate value from existing digital archives. Moreover, mobile scanning captures notes and field documents, allowing frontline staff to contribute content to the corporate knowledge base. Therefore, organizations can boost AI readiness by improving content quality without rebuilding repositories.
Bochniak walks through admin steps that show OCR setup in the Microsoft 365 admin center and links to billing via Azure or the pay-as-you-go model. As a result, administrators must decide whether OCR should run across all sites by default or be limited to specific locations, and that decision affects cost and governance. Additionally, proper configuration helps prevent unnecessary processing and keeps sensitive content from being indexed unintentionally.
However, enabling OCR everywhere can increase costs and surface low-quality text that dilutes search results, so organizations face tradeoffs between broad coverage and targeted processing. Therefore, estimating usage and running pilots on high-value sites helps balance cost against expected benefits. In short, governance and cost management are essential when rolling out OCR at scale.
Despite its advantages, OCR does not perfectly capture every document, and accuracy varies with handwriting, image quality, and language support. Consequently, organizations should expect some errors, formatting loss, and occasional false positives that can create noise in search results and AI outputs. Moreover, scanning sensitive documents raises privacy and compliance questions that require retention, redaction, and access controls to be in place.
Therefore, Bochniak recommends a pragmatic rollout: pilot OCR on high-value content, monitor extraction quality, and apply human review where accuracy matters most. In addition, teams should tune site-level settings and retention policies to limit exposure and control costs. Ultimately, the video presents OCR as a practical, cost-manageable step to improve data quality for Copilot, while also reminding organizations to balance accuracy, privacy, and budget when adopting the feature.
In summary, the video by Szymon Bochniak (365 atWork) offers a clear, step-by-step look at how to use free OCR in Microsoft 365 and the Copilot mobile app to make image-based documents usable for AI and search. Consequently, teams can unlock useful information from legacy archives and field scans without major infrastructure changes. However, leaders must weigh the tradeoffs between scope, cost, accuracy, and compliance before enabling OCR broadly.
Finally, the recommendation is to start small, measure impact, and expand coverage where the business case is strongest, thereby improving Copilot’s usefulness while controlling risk. In this way, the approach enables organizations to make faster, better decisions based on content that used to be invisible, and it does so with manageable steps and clear governance considerations.
microsoft 365 ocr, free ocr microsoft 365, copilot data quality, improve copilot data, ocr for copilot, document scanning microsoft 365, enhance data quality with ocr, optimize copilot performance