StackDC AI OCR

Sign in to StackDC, upload a document, and extract structured data with AI reasoning.

StackDC AI OCR streamlines the process of turning document content into structured, reusable data inside your browser. Built as a side-panel extension, it is designed for users who regularly move information from PDFs or image files into online forms and data-entry workflows. Instead of manually retyping values, the extension combines document upload, AI-powered extraction, field preparation, result review, and form filling in one guided experience. The extension supports signing in to a StackDC server and working directly from the current browser tab. Users can define which fields should be extracted, upload one or more documents, run the extraction process, and review the structured output before applying it. This makes the workflow especially useful for repetitive business tasks where information from scanned documents or digital files must be transferred accurately into web-based systems. One of its most practical capabilities is on-demand page scanning. When asked by the user, the extension can inspect the active tab and identify visible form controls that are suitable for document-driven completion. This includes text inputs, textareas, dropdowns, radio groups, checkboxes, contenteditable areas, and supported ARIA-based editable widgets. Those detected form fields can then be turned into extraction targets automatically, reducing setup time and helping users build a field list from the page they are already working with. The side panel includes both a fast workflow and a more advanced setup mode. In the quick workflow, users can scan the current page, upload supporting documents, extract answers, and fill the form in a streamlined sequence. In advanced mode, users have more control over field creation and document handling. They can define field names, add optional descriptions, specify data types, manage options, and configure unique-key behavior. The extension also supports importing and exporting field templates in Excel format, making it easier to reuse extraction setups across repeated tasks. After extraction, the results can be reviewed and corrected before they are used. The package shows that edited answers become the values used for later actions such as export, form filling, and saving to a database. The matching process between extracted answers and page fields is handled carefully through multiple matching passes, with an emphasis on avoiding ambiguous assignments where possible. This helps keep the workflow controlled and makes the extension more suitable for practical form completion tasks. The package also points to template-based and database-oriented workflows. Users can load saved table definitions, compare fields against existing tables, preview schema changes, and save extracted answers to database tables through StackDC API endpoints. Export options are available for templates and extraction results using spreadsheet files, helping users move structured data into broader operational processes. Additional interface elements such as progress displays, document preview controls, workflow step indicators, and sign-in state handling contribute to a guided experience from upload through review and final use. A notable detail is the separation between scanning and filling. The extension scans the page only when the user explicitly requests it, and filling is limited to controls found during the latest scan. The scanner also avoids hidden, disabled, password-related, credential-related, and payment-card-related inputs, focusing instead on visible and editable fields relevant to document-based completion. Key Features: - AI-powered extraction from PDF and image documents into structured answers based on selected fields. - Side-panel workflow that combines sign-in, upload, extraction, review, and filling in one interface. - Quick Fill process for scanning the current tab, extracting data from documents, and filling detected form fields. - Advanced field setup with support for field names, descriptions, data types, options, and unique-key settings. - On-demand form scanning for visible inputs, textareas, selects, radio groups, checkboxes, contenteditable areas, and supported ARIA widgets. - Excel import and export for templates, plus spreadsheet export for extraction results. - Saved templates and database table support, including table loading, field comparison, schema preview, and database save actions. How It Works: 1. Open StackDC AI OCR from the browser toolbar to launch the side panel. 2. Sign in to your StackDC environment and configure the server URL if needed. 3. Upload one or more PDF or image documents, or start from the current page using the quick workflow. 4. Create extraction fields manually, import them from Excel, load a saved table template, or scan the current tab to build fields from detected form controls. 5. Run the extraction process so the selected documents and field configuration are sent to the StackDC service for structured result generation. 6. Review the extracted answers, make corrections if needed, and optionally export the results or save them to a database table. 7. Fill the active page with the reviewed values using the fields detected during the most recent page scan. Privacy: - This extension collects browsing data to improve user experience. The package indicates that it can scan the active tab for form fields when requested by the user, upload selected documents through the configured StackDC API for extraction, and store authentication tokens in session storage. It may also remember the username and an encrypted secret on the device when the remember option is used, with local storage and IndexedDB-based key handling. The selected server URL and certain interface preferences are also stored locally.