What is esumry?

esumry is a comprehensive AI-powered legal case management platform that dramatically reduces the time, effort, and cost of managing legal cases and reviewing documents. It provides a complete case management system with AI-powered document analysis, interactive chat sessions for case research, and comprehensive workflow tools. esumry assists you with case organization, document processing, AI-powered analysis, and generating insights for documents, pleadings, contracts, and other legal materials. The platform is powered by advanced AI models from OpenAI's series 4 and 5 model families.

How do I use esumry?

Access esumry with your internet browser and use it from anywhere you have internet service. Start by creating a new case to organize your legal matter, then upload documents through our streamlined workflow. The platform provides a complete case management system where you can organize documents, take notes, and use AI-powered chat sessions to analyze your materials. Ask questions in natural language about your case documents and receive intelligent responses with citations. You can also generate comprehensive summaries, organize case information, and review case strategy all within the platform.

How does the case management system work?

The esumry case management system provides a complete workflow for organizing your legal matters. Create new cases with detailed information including case type (from antitrust to personal injury), contentions, and descriptions. Each case provides a central dashboard where you can access case notes, manage documents, and organize chat sessions. The system supports different case types and provides tools for tracking case progress, organizing documents, and maintaining comprehensive case records. You can easily switch between multiple cases and access everything from a unified dashboard.

What can I do with AI chat sessions?

AI chat sessions let you have intelligent conversations about your case documents in natural language. Responses stream in real time as the AI generates them, and numbered citations (e.g., [1], [2]) in each response are clickable links — clicking a citation opens the Citation Explorer, which shows the exact source text on the left and the surrounding document excerpt with page number on the right.

Agentic processing: When you send a message, esumry runs an agentic workflow. The assistant searches your scoped documents with tools, can perform follow-up reads across multiple steps, and then synthesizes a final cited answer.

You can scope chat sessions to focus the AI's search on a specific subset of your documents:

• Full case: The AI searches across all documents in the case.
• Category-scoped: Limit the session to one document category (Transcripts, Pleadings, Medical Records, Exhibits, Key Documents, or Discovery Documents).
• Tag-scoped: Limit the session to documents attached to a specific user-created tag, regardless of category.

You can create multiple chat sessions for different research topics, and the AI maintains conversation context so you can follow up with additional questions.

Medical Records sessions: Category-scoped chat sessions limited to Medical Records can supplement answers from your uploaded records with authoritative reference lookups from MedlinePlus and RxNorm (U.S. National Library of Medicine) and openFDA (drug labels, adverse events, devices, and recalls). See What reference tools are available in Medical Records chat sessions? below.

Session management: Use the three-dot options menu in the chat header to rename or delete a session, or to download the session as a Word document (.docx).

What reference tools are available in Medical Records chat sessions?

When you start or open a Medical Records category-scoped chat session, the assistant still searches your uploaded medical records first and cites them with numbered references (e.g., [1], [2]) that open in the Citation Explorer. In addition, the agent can pull general medical reference data from trusted public sources to help explain terminology, medications, and regulatory context mentioned in your records or in your question.

Available reference sources:

• MedlinePlus (National Library of Medicine) — health topic summaries for conditions, procedures, anatomy, and related concepts
• RxNorm (National Library of Medicine) — normalized medication names, strengths, dose forms, and RxNorm identifiers
• openFDA — FDA drug labeling (indications, warnings, interactions, dosage), voluntary drug adverse event reports (FAERS), medical device classification and adverse events (MAUDE), and drug or device recall enforcement reports

These lookups supplement your case documents; they do not replace citations to the medical records themselves. Answers label external references distinctly in the response text (for example, [MedlinePlus Topic (Asthma)], [RxNorm Concept (...)], or [FDA Drug Label (...)]). MedlinePlus-backed answers also include a MedlinePlus references: section with links to the topic pages.

After the assistant searches your medical records for a question, esumry may automatically run relevant MedlinePlus, RxNorm, and FDA drug-label lookups when your question calls for them. The assistant can also request additional openFDA lookups (such as adverse events, devices, or recalls) when that context would help answer your question.

The medical records session view includes an info note at the bottom of the page listing these tools. Reference data is for legal case analysis support only and is not medical advice.

How do I upload and organize documents?

Upload documents easily using our enhanced drag-and-drop interface or file selector. You can upload up to 10 documents simultaneously with our new multi-document upload feature, each up to 20MB. For batch processing, you can also upload ZIP files that hold multiple documents - the system automatically extracts and processes each file individually. Documents are automatically associated with your cases and processed through our AI pipeline for text extraction, entity identification, and search indexing. Monitor processing status in real-time with improved batch processing capabilities and begin analysis as soon as processing completes. The system provides clear feedback about processing progress on each document for efficient case management.

Can I summarize medical records?

Yes. Medical records support both Short and Long AI-generated summaries.

• Short summary: A concise chronological narrative covering diagnoses, treatments, procedures, medications, and physician opinions with page citations.
• Long summary: A comprehensive, detailed narrative that preserves every clinically relevant detail.

Medical records also feed the Case Timeline. After processing, esumry extracts dated clinical events and can attach structured ICD-10 codes, procedures, medications, and a documented time of day to each timeline event when that information appears in the record. See How does the case timeline work? for how to view, edit, and export those details.

In Medical Records chat sessions, the assistant can also use MedlinePlus, RxNorm, and openFDA reference tools alongside your uploaded records. See What reference tools are available in Medical Records chat sessions?

You can also ask specific questions about a medical record or summarize sections of it through the Chat function.

When your case has two or more medical records with completed neutral Short or Long summaries, you can request a Consolidated Medical Record Summary — one case-level narrative that merges every medical record into a single chronological story. See What is the Consolidated Medical Record Summary?

What is the Consolidated Medical Record Summary?

The Consolidated Medical Record Summary (Beta) is a single AI-generated narrative that merges every medical record on a case into one chronological, litigation-ready story. Instead of reading each record's Short or Long summary separately, you get one case-wide overview with citations back to the source medical records.

How to request it:

• On the Medical Records list, click Request Consolidated Summary next to Start Chat when the case is eligible
• From the Summary Catalog, select the Consolidated row and click Request (or include it in a bulk Request Selected action)

Before you can request:

• The case must have at least two medical record documents
• Every medical record must have a completed neutral Short or Long AI summary
• No medical record summaries may still be generating

Consolidated summary viewer page: Open View Consolidated Summary from the medical records list or Summary Catalog. The page includes:

• View / Edit tabs — read the formatted narrative or edit the raw text; use Hide citations on the View tab to read without bracketed references
• Cross-document citations — citations name the source medical record and link to that record's PDF viewer at the cited page
• Index and Ask AI — browse important terms with line links and ask questions about the consolidated summary
• View source medical records — see which records were merged, with summary variant, file size, and page count
• Save / Auto save, Download (Word or PDF, with optional citation stripping and summary index), and Expand to full page
• Rebuild summary (footer) — queue a fresh consolidated summary when the current one is at least 24 hours old

The consolidated summary is separate from per-record Short and Long summaries — requesting or rebuilding it does not change individual medical record summaries. AI-generated content must be reviewed for accuracy.

How do I manage document metadata?

esumry provides comprehensive document metadata management through a dedicated full-page interface. You can access document metadata either from case details pages or through the document workflow. The metadata page has two tabs:

• View Tab: Displays all AI-generated metadata fields — deponent name, questioner name, court system, other involved parties, locations, topics discussed, contradictions, crucial testimony, objections, crux point, and dates — in a formatted, read-only layout. Citation references (e.g., [4:2-35]) are clickable links that jump directly to the referenced page in the document viewer.
• Edit Tab: Edit any metadata field directly, with changes saved automatically as you type.

The AI automatically identifies key objections raised during depositions, hearings, and trial transcripts, capturing the objecting party, type of objection, and a page/line citation link for each entry.

Use the Download button on the metadata page to export transcript metadata as a Word (.docx) or PDF file. A field selection dialog lets you choose exactly which fields to include, with Select All and Deselect All shortcuts. If the transcript belongs to a case, you can also check Include all case transcripts to generate a single combined document covering every transcript in the case.

All metadata is automatically integrated with your case navigation and helps enhance AI-powered analysis and search capabilities.

How do you manage security and privacy?

We are SOC 2 Type I certified and take data security and privacy very seriously. Your data is encrypted using enterprise best practices for data at rest and in transit. Documents are not available except to the authenticated account user. esumry uses models from OpenAI's series 4 and 5 model families for automated summarization, agentic case chat, and related AI features. Open AI does not use the data it processes for service improvements and does not retain the data after it is processed (zero data retention). See our Terms and Conditions for details.

Our information security policy is available upon request.

How soon can I get started?

You can start as soon as you are notified that your account is approved, usually within one business day of signing up for esumry. Once approved, sign in with your username and password, or use Sign in with Microsoft when that option is available on the login page.

Can I annotate the transcript on my screen?

Yes. Open any document in the PDF viewer — for example by clicking a citation link in a summary or metadata field, or choosing View from a document list. Click Expand to full page in the toolbar to hide the app navbar, sidebar, and footer for a distraction-free reading view; press Escape or click the button again to exit. Click Index in the toolbar to browse important terms, click Timeline when dated events have been extracted to review this document's chronology and jump to cited pages, click Citations on pleadings to open the External Citations panel (Beta) when citations have been extracted, or click Annotate to reveal markup tools.

You can mark up the PDF directly on any page:

• Highlight — drag to highlight a region of text
• Pen — draw freehand notes
• Text — click on the page to place a text callout
• Area box — draw a rectangular box around a section
• Eraser — remove pen strokes

Use the Select tool (cursor icon) to click an existing annotation, then remove it with the trash button or the Delete key. Annotations save automatically as you work. A Saving / Saved indicator next to the Annotate button confirms your changes are stored, and your markup reloads the next time you open the document.

Download with annotations: Click the Download button (download icon) in the PDF viewer toolbar to save a copy of the document with your annotations permanently burned into the PDF pages. esumry saves any pending annotation changes first, then generates the file — a progress modal appears while the PDF is being prepared. The downloaded file is named with an _annotated suffix (for example, deposition_annotated.pdf) and includes highlights, pen marks, text callouts, and area boxes. Use it for sharing markup with colleagues or offline review outside esumry.

To edit the AI-generated summary text itself, use the Edit tab on the Summary page — that is separate from PDF viewer annotations.

Does it work with Microsoft Word®?

Yes. esumry can export your work as Microsoft Word (.docx) or PDF files from several places in the platform:

• Document Word/PDF File: Available from the Document Digest page. A content selection modal lets you choose what to include: document crux point metadata, AI-generated summary, Q&A extraction table, and detailed document metadata. Choose between Word (.docx) or PDF output format. Citations in the summary or metadata text (e.g., [4:2-35]) become clickable hyperlinks in the downloaded file, linking to the referenced page in the appended source document.
• Summary Word/PDF File: Available from the Summary page. Downloads the document's AI-generated summary (including any edits you've made) as a Word (.docx) or PDF file. Citations in the summary (e.g., [4:2-35]) become clickable hyperlinks that jump to the referenced page in the appended source document. When a summary index exists, check Include summary index in the download dialog to append it to the file.
• Chat Session Word/PDF File: Download any chat session from the session options menu (three-dot menu in the chat header). Choose between Word (.docx) or PDF output format.
• Transcript Metadata Word/PDF File: Download transcript metadata from the metadata page. A field selection dialog lets you choose exactly which fields to include. Choose between Word (.docx) or PDF output format. Citations in metadata fields (e.g., [4:2-35]) become clickable hyperlinks in the downloaded file, linking to the referenced page in the appended source document.
• Case Timeline: Export as a Word document or Excel spreadsheet from the Timeline page Download button, with options to include all events or only events from selected documents, and to group events by document or by date. Exports include ICD-10 codes, procedures, medications, and event time when those fields are present on medical record events.
• Draft Answer Word File: Download from the Draft Answer page, with the choice to include or exclude the original complaint paragraphs.
• Pleading Citations Report (Word/PDF): From the PDF viewer Citations panel (Beta) on pleadings, download a report of extracted citations. Choose Word (.docx) or PDF format and optionally include clickable in-document links in the export.
• Annotated PDF from PDF viewer: Click the Download button (download icon) in the PDF viewer toolbar to save a PDF copy with your annotations burned into the pages. The file is named with an _annotated suffix and includes highlights, pen marks, text callouts, and area boxes.
• Consolidated Medical Record Summary (Word/PDF): Download the case-level merged medical record narrative from the consolidated summary viewer page. Choose Word (.docx) or PDF, optionally exclude citations, and include the summary index when available.

How does the AI-powered case analysis work?

The esumry AI system provides comprehensive legal document analysis through interactive chat sessions and automated document processing. You can ask questions about your case documents in natural language and receive intelligent, cited responses. The AI understands legal context, can analyze multiple documents simultaneously, identify key entities, and provide insights across your entire case collection.

Chat sessions can be scoped to focus the AI's search on a specific subset of your documents:

• Category-specific sessions: Limit the AI to documents of a single category — Transcripts, Pleadings, Medical Records, Exhibits, Key Documents, or Discovery Documents.
• Tag-specific sessions: Limit the AI to documents attached to a user-created tag, which can span multiple categories.

Every AI response includes numbered citations (e.g., [1], [2]). Clicking a citation opens the Citation Explorer, which shows the exact source text the AI used, the surrounding document excerpt with page number, and a direct link to that location in the full document viewer. The system maintains context throughout your conversations.

How do case notes work?

The case notes feature provides a dedicated environment for capturing and organizing your added case-related information. Click Expand to full page in the toolbar to hide the app navbar, sidebar, and other chrome for a distraction-free writing view. Notes are automatically saved as you type, and you can format text using rich text editing options. Use case notes to record client meetings, legal research findings, track case strategy, maintain case timelines, and prepare for depositions or court appearances. All notes are associated with specific cases and easily accessible from your case dashboard.

Can I manage multiple cases at once?

Yes, esumry is designed for managing multiple cases efficiently. The Cases dashboard provides an overview of all your cases, and the Cases List offers advanced search and filtering capabilities. You can quickly switch between cases, organize them by different criteria, and access recent activity across all your cases. Each case organizes its own documents, notes, and chat sessions while providing centralized access from the main dashboard.

How long does it take to process documents?

Document processing time depends on file size and complexity. Document analysis usually completes within a few minutes through our AI processing pipeline, which includes OCR text extraction, document structure analysis, entity identification, and search index generation. You can track processing status in real-time and begin using AI chat analysis as soon as processing is complete. Large document collections may take longer, but esumry can process multiple documents simultaneously, so you can start working on completed documents while others are still processing.

Do you store my documents and case information?

We securely store your case documents, AI-generated summaries, case notes, and chat conversations to provide you with a persistent, organized case management experience. All data is encrypted using enterprise-grade security practices for data at rest and in transit. You maintain full control over your data - you can delete individual documents, entire cases, or your entire account at any time. Once deleted, data is permanently removed from our servers and cannot be recovered. We also provide detailed purchase history and account management tools, so you always know what information we have about your account.

Which file formats does esumry support?

esumry supports a wide range of legal document formats including:

• PDF documents - preferred format for optimal processing
• Microsoft Word documents (.doc, .docx)
• Plain text files (.txt) - ideal for documents and transcripts
• ZIP archives (.zip) - for batch uploading multiple documents at once
• Image files (.jpg, .jpeg, .png, .webp, .tiff, .heic) - supported for Exhibits, Key Documents, and Discovery Documents

The platform can handle documents, contracts, pleadings, expert reports, and other legal documents up to 20MB each. Advanced OCR technology extracts text from scanned documents, and the AI processing pipeline optimizes content for legal analysis and search.

ZIP File Uploads: When uploading a ZIP file, the system automatically extracts all supported document types (.txt, .pdf, .doc, .docx) and processes them individually. Each file within the ZIP must meet the 20MB size limit. Nested ZIP files are automatically skipped to prevent processing issues. The total number of files (including those extracted from ZIP archives) is limited to 10 files per upload batch.

Do you support condensed or mini-script PDFs?

No, esumry does not support "condensed" or "mini-script" PDF formats. These are special PDFs where four pages of a document are displayed in quadrants on a single PDF page (one page in each corner). While this format is sometimes used to reduce printing costs, it creates significant challenges for accurate text extraction and document processing.

How to identify condensed PDFs:

• Each PDF page shows four miniature document pages arranged in a 2x2 grid.
• Text appears very small and may be difficult to read without zooming.
• Page numbers show four different numbers per PDF page.
• The document appears to have 1/4 the number of pages it should have.

What to do if you have a condensed PDF:

• Request a standard format: Contact your court reporter or document provider and request a standard single-page-per-PDF format.
• Alternative formats: Ask for the document in .txt (plain text) format, which is ideal for depositions and transcripts.
• Microsoft Word format: Request .doc or .docx format as an alternative to PDF.

Note: Most court reporters can provide transcripts in standard formats, but you may need to specifically request it. Standard-format documents ensure optimal processing, accurate text extraction, and the best experience with the esumry AI-powered analysis features.

How do I manage my account and subscription?

Your Account Profile page is the gateway to managing all aspects of your account. On this page, you can update your personal and professional information, manage subscription changes, change your password, and request account deactivation if needed. From here, you can also use our secure Stripe integration to view your complete purchase history, update payment methods, and access detailed invoices for accounting purposes.

On the login page, you can sign in with your username and password, or — when enabled for your environment — use Sign in with Microsoft for Microsoft Entra ID single sign-on. See Can I sign in with Microsoft? for details. Our account management system provides full transparency and control over your esumry experience.

Can I sign in with Microsoft?

Yes, when Microsoft single sign-on (SSO) is enabled for your esumry environment. On the login page, click Sign in with Microsoft to authenticate with your work or school Microsoft account through Microsoft Entra ID.

• You must already have an approved, active esumry account with a verified email.
• Your Microsoft account email must match the email on your esumry account (matching is case-insensitive).
• SSO signs in existing users only — it does not create a new esumry account.
• Username and password login remains available alongside SSO.

If you do not see the Microsoft button, SSO may not be enabled for your environment. If sign-in fails, confirm your email is verified and matches your Microsoft account, or contact support@esumry.com.

I found a bug. How do I report it?

Please send an email to support@esumry.com. Please include as much detail as possible, like what you were doing when the bug happened and any error codes that may have been displayed.

How can I request new features?

Please send an email to support@esumry.com.

How are documents organized into categories?

Every document uploaded to esumry belongs to one of six categories. Categories determine the Case Details page section in which a document appears and can be used to scope AI chat sessions to only that document type.

• Transcripts — depositions, hearing transcripts, trial transcripts
• Pleadings — complaints, answers, other pleadings
• Exhibits — trial and deposition exhibits
• Medical Records — processed with enhanced OCR for handwritten and scanned medical documents; timeline date extraction can attach ICD-10 codes, procedures, medications, and event times to dated clinical events
• Key Documents — important reference materials for the case
• Discovery Documents — discovery-related materials

You select a category when uploading. If you want to chat with documents across multiple categories together, create a Tag and attach documents from any combination of categories to it. On the Case Timeline page, categories also power the right sidebar By Type filter, which lets you quickly include or exclude documents by category.

How do case tags work?

Tags let you create custom, cross-category groupings of documents within a case. Use them to organize documents by topic, witness, or any other criteria relevant to your matter.

• Create up to 5 tags per case, each with a custom name, description, and color.
• Attach any document to a tag regardless of its category. A deposition, a medical record, and an exhibit can all be attached to the same tag.
• The same document can belong to multiple tags.
• Start a tag-specific AI chat session that queries only the documents assigned to that tag.
• Manage tags from the Tags page for each case — Add, edit, delete, rename, or recolor tags; manage which documents are attached from the Attach Documents page.

Tags are optional. If you do not need custom groupings, document categories alone are sufficient for organizing your case.

How does the case timeline work?

The Case Timeline is an interactive chronological display of important dates and events drawn from your case documents.

• Three views: Switch with the Chart, List, and Table toggle in the timeline header — Chart (interactive visual graph), List (chronological text listing, with optional grouping by document), and Table (sortable columns with pagination, aligned with timeline Excel export fields)
• Table view: Review events in a spreadsheet-style table. Sort by clicking column headers (document, category, dates, title, description, locations, individuals, citations, and more). When your document filter includes Medical Records, columns for ICD-10 codes, procedures, and medications appear as well. Search filters events by title or description across the full filtered set; results paginate 25 events per page. Use the edit action on a row to update an event. Ask AI works on the events currently displayed, including after search, with saved chat history.
• AI Extraction: The AI automatically finds dates and events, then extracts them from uploaded documents — including locations and individuals (each with a name and role) when mentioned in the source text. Medical Records use a dedicated extraction pipeline that can also attach ICD-10 codes, procedures, medications, and a documented time of day to each dated event when that information appears in the record.
• Event Details: Each event displays its title, description, date(s), source document, and clickable citation links. When available, events also show locations, individuals (with roles), and — for medical record events — ICD-10 codes, procedures, medications, and an event time (shown with the date on List view when documented). Edit any event to update these fields.
• Medical Record Event Fields: When you create or edit a timeline event tied to a Medical Records document, the modal includes separate tabs for ICD-10, Procedures, and Medications. Each ICD-10 entry has a code and description; procedures and medications use a name plus optional details (for example, CPT/HCPCS codes for procedures or dosage and route for medications). You can add, edit, or remove rows in each tab.
• Manual Events: Add your own custom events at any time. Document, Start Date, Title, and Description are required. End Date is optional for events that span a range. You can also add optional locations, individuals (name and role), and — for medical record documents — clinical fields in the ICD-10, Procedures, and Medications tabs.
• Navigation (Chart view): Jump forward and backward, event-by-event, using Previous/Next buttons. Zoom in and out to focus on a date range. Show All Events to reset the view.
• Filters: Filter by date range or by specific documents; events are color-coded by source document. In the right sidebar Documents section, use individual checkboxes or Toggle All, or click By Type to open a Document types modal and quickly select documents by category (Transcripts, Pleadings, Medical Records, and so on). Choose one or more types and click Apply to check only timeline-ready documents in those categories; choose All types to clear the type selection before applying.
• Citation Links: Click any timeline event to view the source document citation.
• Export: Download as a Word document or Excel spreadsheet (or both), with options to include all events or only events from selected documents, and group events by document or by date. Exports include locations, individuals, citations, and — when present — ICD-10 codes, procedures, medications, and event time for medical record events.
• Ask AI (List and Table views): Click Ask AI in the toolbar to ask questions about the events currently displayed (in Table view, this includes search-filtered results). Conversations are saved automatically and can be resumed from Chat history. Chart view and the List view filter sidebar also have Ask AI for filtering events by date range or documents using natural language; those filter chats are not saved.

Access the timeline from the Case Options menu (three-dot menu) on the case overview page.

While viewing a single document in the PDF viewer, open the document-level Timeline panel to review that document's dated events and jump to cited pages without leaving the viewer. See What is the Timeline panel on the PDF viewer?

What is the Draft Answer feature?

The Draft Answer page helps you draft responses to complaint documents, paragraph by paragraph, with AI assistance at each step.

• Upload a complaint to a case and open Draft Answer from the pleading's actions menu once paragraph extraction has finished.
• Each complaint paragraph is displayed in its own card with an editable answer text area below it.
• Answer Options: Choose from Deny, Lacks Info (defendant lacks sufficient information to admit or deny), Add Context, or Admit. The AI generates a tailored response based on your choice.
• Interview Mode: Start Interview walks you through each paragraph one at a time, prompting you to provide factual context before the AI drafts a response. This produces more accurate, case-specific answers than one-click generation.
• Client Questionnaire: Click Questionnaire to generate a separate set of Word documents for your client. Confirm the linked case type (required), then click Build Questionnaire. esumry analyzes the complaint, identifies defendants, and drafts fact-gathering questions tailored to the allegations. When generation completes, download a ZIP file containing one questionnaire .docx per defendant (shared questions plus defendant-specific questions). Each document includes numbered questions with space for client answers and an AI disclaimer. Use Retry to regenerate if needed. This export is independent of the draft answer Word download on the page footer.
• Affirmative Defenses: Add and edit affirmative defenses below the numbered paragraphs; they are included in the draft answer export.
• Progress Tracking: The paragraph index on the left highlights completed paragraphs in green so you can see your progress at a glance.
• Auto-save: Answer text areas save automatically as you type.
• Ask AI: Use the purple Ask AI sidebar for conversational help drafting or updating answers (saved chats supported).
• Download: Export the completed draft answer as a Word document from the page footer, with the option to include or exclude the original complaint paragraphs.

What is the Contentions page?

When a case has both a complaint and an answer on file and both have finished processing, esumry builds a Contentions page from those pleadings. Open it from the case options menu (three-dot menu) on the case overview page.

For each contention, the page shows:

• A plain-language summary of the dispute issue
• The matching complaint paragraph
• The corresponding answer response

Affirmative defenses from the answer are listed in a separate section at the bottom of the page.

You can set importance (high, medium, or low) on each contention and affirmative defense to reflect how central it is to the case. Only high and medium importance items are used when generating plaintiff- or defendant-focused Short summaries for transcripts. Changes save automatically.

The page toolbar also includes Expand to full page, which hides the app navbar, sidebar, footer, case header, and section navigation so you can focus on the contentions list. Press Escape or click the button again to exit.

Use the Ask AI button on the Contentions page to ask questions about the case contentions in a context-aware sidebar.

How do Q&A pairs and the extraction table work?

The Document Digest page is the primary interface for reviewing transcripts and building extraction tables.

• Select Text: Click and drag to select any passage in the document or use the page-select button to select an entire page. When text is selected, two options appear: Copy Selection (adds the text as-is to the extraction table) or Use AI to Summarize Selection (generates an AI summary of the selected passage).
• Q&A Pairs: Toggle Show QA Borders to display colored borders around question-and-answer exchanges in the transcript, making them visually distinct. Check individual Q&A pairs to include them in the extraction table.
• Copy All QA Items: Use the Copy All QA Items button to batch-add all currently visible Q&A pairs to the extraction table at once, skipping any duplicates.
• Extraction Table: The right sidebar displays all extracted items — location, excerpt, and notes. Items can be edited or deleted individually.
• Delete All: The More Actions menu (three-dot dropdown) includes a Delete All Extraction Items option that clears the entire table after confirmation.
• Document Index: The sidebar also contains a document index for quick navigation to sections of the transcript.
• Word Cloud: Open a word cloud from the More Actions menu (three-dot button) on the Document Digest page, or from the toolbar in the document viewer. The word cloud displays the most frequently occurring terms in the document, with word size reflecting how often each term appears. Common filler words and terms that appear on every page are automatically filtered out.
• Sidebar Toggle: Use the sidebar toggle button to hide the right panel and maximize the document viewing area.
• Download: The extraction table content is optionally included in the Word document download via the content selection modal.

How do I search case transcripts from the Document Digest?

When a transcript on the Document Digest page belongs to a case, a search icon appears in the bottom toolbar (to the left of Show QA Borders). Click it to open the Search case transcripts modal and search across every transcript in that case without leaving the digest you are reviewing.

Enter keywords and click Search. Matching passages are grouped by transcript in expandable sections. Each result shows the page or page/line location, the matching excerpt, and a View Page link that opens that transcript at the cited page.

Use the settings gear next to the Search button to adjust:

• Total results — how many matches are returned across all transcripts
• Results per document — how many matches are shown per transcript

Your slider preferences are remembered for future searches. When results are displayed, use the Download button in the modal footer to export them as a Word (.docx) file.

Search uses the same document search engine as Case Chat and becomes available once every transcript in the case has finished indexing for chat. If indexing is still in progress, the modal shows a waiting message and checks automatically until search is ready.

How do I view and edit document summaries?

The Summary page for each document has two tabs:

• View Tab: Displays the formatted AI-generated summary. Citation references (e.g., [4:2-35]) are clickable links — click any citation to jump directly to that page in the document viewer. Use this tab to read and navigate the summary.
• Edit Tab: Change the summary text directly. Use this tab to correct the AI output, add your own analysis, or rewrite sections. Changes are saved and immediately reflected in the View tab.

Click Expand to full page in the page header to hide the app navbar, sidebar, and page chrome. Press Escape or click the button again to exit.

When an index is available, click Index in the page header to open the index panel.

Additional summary actions:

• Remove Citations: Available in the More Actions menu (three-dot button), this permanently strips all bracketed citation references from the saved summary text. A confirmation modal shows how many citations will be removed before applying. This is different from Do not include citations on download, which affects only the exported file.
• Feedback: Use the thumbs up / thumbs down buttons to rate the summary and optionally include a feedback message to help improve the AI.
• Save and Download: Save your edits, then click Download to export as Word (.docx) or PDF. The download dialog lets you choose the format and citation handling: check Do not include citations to remove citation references from the file only (your saved summary is unchanged), Include clickable citations to append the source document and make citations clickable links in the export, or Include summary index when an index is available to append it to the file. Your preferences are remembered for future downloads.

Short and Long summaries: Each document can have both a Short summary (quick overview) and a Long summary (comprehensive narrative). Clicking "View" on a document list opens a selection modal where you can choose which summary to view or trigger generation of a missing variant. On the summary page, a dropdown lets you switch between available summaries. In the case summaries sidebar, a second dropdown shows all variants for the selected document.

To review every AI summary across the case — what's ready, missing, or processing — and request many at once, open the Summary Catalog. See What is the Summary Catalog?

Choosing summary length at upload: When uploading a document, you can select Short, Long, or None (skip summary generation). You can generate or regenerate either variant from the summary page at any time.

What are point-of-view (POV) short summaries for transcripts?

Transcript documents — depositions, hearings, and trial transcripts — can have Short summaries written from a Neutral, Plaintiff, or Defendant point of view. Long summaries remain a single neutral, detailed narrative.

When uploading a transcript, choose Short as the summary length, then select a perspective under Summary POV:

• Neutral — a balanced overview of the testimony
• Plaintiff — organizes testimony around issues from your case Contentions page from the plaintiff's perspective
• Defendant — organizes testimony around those same issues from the defendant's perspective

Plaintiff- and defendant-focused Short summaries use the case Contentions page. Set importance (high, medium, or low) on each contention and affirmative defense there; only high and medium importance items are included when generating those summaries.

Summary POV is available for transcripts only — not for pleadings, exhibits, medical records, Key Documents, or Discovery Documents. On the document summary page, use the summary dropdown to switch between available Short and Long variants, including different POVs when they have been generated.

What is the Summary Catalog?

The Summary Catalog is a case-level page that lists every AI summary variant for your documents in one table — so you can see what is ready, still missing, or processing, and request many summaries at once.

How to open it:

• From any case page, open the case options menu (⋯) and choose Summary Catalog
• From a document's Select Summary Type modal (when you click View on a summarizable document), choose View Summary Catalog
• From the document Summary page, use the summary variant dropdown and select Summary Catalog

What you can do:

• See one row per applicable AI summary slot (Short and Long Neutral; Short Plaintiff and Short Defense for transcripts), plus a Consolidated row (Beta) when the case has medical records — request or view the case-level consolidated medical record summary from that row
• Filter by document name, category, summary type, or status (Available, Missing, or Blocked)
• Select missing or failed summaries and click Request Selected to queue generation in bulk (up to 50 at a time)
• Click View on a ready summary to open it on the document Summary page
• Review Total Reading Time Saved for completed summaries in your current filter
• Use Ask AI for help deciding which summaries to request (saved chat history supported)

What can I download or export?

You have several export options to save your work outside the platform:

• Document File (.docx or PDF): On the Document Digest page, a content selection modal lets you choose what to include: document crux point metadata, AI-generated summary, Q&A extraction table, and comprehensive document metadata (deponent, questioner, dates, locations, contradictions, crucial testimony). Choose between Word (.docx) or PDF output format. Citations in the summary text (e.g., [4:2-35]) are converted to clickable hyperlinks in the downloaded file that jump directly to the referenced page in the appended source document.
• Document Summary File (.docx or PDF): On the Summary page, download the AI-generated summary (with any edits applied) as Word (.docx) or PDF. In the download dialog, choose Do not include citations for a clean export without bracketed references, Include clickable citations to append the source document and convert citations (e.g., [4:2-35]) into hyperlinks that jump to the referenced page, or Include summary index when an index is available to append it to the export.
• Chat Session File (.docx or PDF): Download any chat session from the session options menu (three-dot menu in the chat header). The download dialog includes an Include Citations option — when checked, AI responses include inline footnote numbers with a citations list below each response showing the source document, page, and line range. Choose between Word (.docx) or PDF output format.
• Transcript Metadata File (.docx or PDF): Download transcript metadata from the metadata page. A field selection dialog lets you choose exactly which fields to include — Deponent Name, Questioner Name, Court System, Other Involved Parties, Locations, Topics Discussed, Contradictions, Crucial Testimony, Objections, Crux Point, and Dates — with Select All and Deselect All shortcuts. For cases with multiple transcripts, checking Include all case transcripts generates a single combined file. Choose between Word (.docx) or PDF output format. Citations in the metadata fields (e.g., [4:2-35]) are converted to clickable hyperlinks in the downloaded file that jump directly to the referenced page in the appended source document.
• Case Timeline Word (.docx) or Excel Spreadsheet (.xlsx) File: Export from the Timeline page Download button. Options include: all events or selected documents only; group events by document or by date. Medical record events can include ICD-10 codes, procedures, medications, and documented times of day in the export.
• Draft Answer Word File (.docx): Download from the Draft Answer page, with the choice to include or exclude the original complaint paragraphs.
• Pleading Citations Report (.docx or PDF): On the PDF viewer Citations panel (Beta) for pleadings, download a report listing every extracted citation. Choose Word (.docx) or PDF format and optionally include clickable in-document links that jump to cited pages in an appended copy of the pleading.
• Annotated PDF from PDF viewer: On the PDF viewer toolbar, click the Download button (download icon) to export the document as a PDF with your viewer annotations burned into the pages. The file is named with an _annotated suffix (for example, deposition_annotated.pdf) and includes highlights, pen strokes, text callouts, and area boxes.
• Consolidated Medical Record Summary (.docx or PDF): Export the case-level merged medical record narrative from the consolidated summary viewer page. The download dialog supports Word or PDF output, optional citation removal, and an optional summary index appendix.

Is there a guided tour of the platform?

Yes. Every major page in esumry has an interactive product tour powered by a step-by-step guide that highlights each feature and explains what it does.

Look for the lightbulb button (floating action button) on any page to launch the tour for that page. Tours are available on:

• Cases page
• Case Details page
• Document Upload page
• Document Digest page
• Summary page
• PDF viewer page
• Summary Catalog page
• Consolidated medical record summary page
• Chat page
• Case Timeline page
• Tags and Attach Documents pages
• Metadata page
• Draft Answer page
• Case Notes page
• Account page

Each tour is tailored to the specific features of that page. Tours remember when you have completed them, but you can relaunch them at any time using the lightbulb button.

What is the Citation Explorer?

The Citation Explorer is a panel that lets you verify the source of any AI-generated response. Every answer from the AI includes numbered citations (e.g., [1], [2]). Click any citation number to open the Citation Explorer, which shows:

• the exact source text the AI used on the left side of the panel
• the surrounding document excerpt with the page number on the right
• a direct link to open that location in the full document viewer

The Citation Explorer is available in all AI chat sessions and is the primary tool for verifying that the AI's responses are grounded in your actual case documents.

What is the Citations panel on the PDF viewer?

The Citations panel (Beta) is a sidebar on the PDF viewer for pleadings. When esumry processes a complaint, answer, or other pleading, it automatically extracts references to outside legal authorities — prior cases, statutes, regulations, court rules, constitutional provisions, and similar citations — and records where each citation appears in the document. Click Citations in the viewer toolbar to review every extracted citation.

What you can do:

• Browse citations with status badges and filter by category (Cases, Statutes, Regulations, Rules) or status
• Click in-document location links (for example, [1:12-16]) to scroll the open PDF to that passage
• Open external source links in a separate tab when available
• Download a citation report as Word (.docx) or PDF, with an option to include clickable in-document links in the export
• Use Ask AI in the panel to ask questions about the citations
• Manually edit standalone citation text with the pencil icon; esumry re-parses your change

What is the Timeline panel on the PDF viewer?

The Timeline panel is a sidebar on the PDF viewer that lists dated events extracted from the current document — for example, key dates from a deposition, medical record visit, or other processed document. The Timeline button appears in the viewer toolbar when the document has timeline-ready dated events in its metadata.

What you can do:

• Browse dated events with title, date, and description preview in a vertical timeline
• Sort Oldest first or Newest first
• Filter events with an exact-match search across titles, descriptions, locations, individuals, citations, and — for medical records — ICD-10 codes, procedures, and medications
• Click a citation link (for example, 1:10-15) to jump the open PDF to that page
• Hover an event for a tooltip with full details, including locations, individuals, and medical record fields when present
• Drag the panel edge to resize it on desktop

This panel shows events from the open document only. For a case-wide chronological view across multiple documents, use the Case Timeline page. See How does the case timeline work?

How does the AI Search feature work on document list pages?

Each case section list page — Transcripts, Sessions, Pleadings, Exhibits, Medical Records, Key Documents, and Discovery Documents — has an AI Search button in the page header. Clicking it opens a slide-in sidebar where you can ask questions about the documents in that list.

Important: AI Search works with document names and metadata only — it does not search the full text content of your documents. To ask questions grounded in document content, use a full AI chat session instead.

• Supports multi-turn conversations — ask follow-up questions and the AI maintains context from earlier in the session.
• Conversations are saved automatically as you chat. Use New chat (plus icon), Chat history (clock icon), and Delete (X icon on each row in the history menu) in the sidebar header to manage saved chats — see What does the Ask AI button do? below for details on these controls.
• Closing the sidebar does not delete saved chats — reopen AI Search and use Chat history to resume a previous conversation.

How do I expand a page to full screen for distraction-free reading?

Several pages include an Expand to full page button in the toolbar (expand icon). Click it to hide the app navbar, sidebar, footer, and other page chrome so the main content fills your screen.

Available on:

• PDF viewer — maximize the document viewing area
• Document summary — focus on the summary text
• Case notes — distraction-free note editing
• Contentions — focus on the contentions list without case navigation chrome
• Consolidated medical record summary — focus on the merged case-wide medical narrative

Press Escape or click Exit full page view (compress icon) to return to the normal layout.

On summary pages, side panels such as Ask AI and Index remain available and adjust to the full-page layout.

How do I use the index panel?

The index panel lists important terms and quick-jump links so you can scan vocabulary and navigate long text faster. It is available on the document Summary page, the PDF viewer, and the Consolidated Medical Record Summary viewer page when index data exists for that view. When downloading a summary as Word or PDF, check Include summary index in the download dialog to append the index to the exported file.

The Document Digest page has a separate document index in its right sidebar for transcript navigation; that sidebar also holds the extraction table and is described in How do Q&A pairs and the extraction table work?

What does the Ask AI button do?

Many pages throughout esumry have an Ask AI button that opens a context-aware slide-in sidebar. The sidebar lets you ask questions in natural language and get AI-generated answers relevant to what you are currently working on.

Ask AI is available on the following pages:

• Case Notes — ask questions or get help with your notes
• Contentions — explore questions about the case contentions
• Case Timeline (Chart view) — use natural language to filter timeline events by date range or documents
• Case Timeline (List view and Table view) — ask questions about the timeline events currently displayed
• Document Summary — ask questions about the document summary
• Consolidated Medical Record Summary — ask questions about the case-level merged medical narrative and its index
• Summary Catalog — get help reviewing summary coverage and deciding what to request
• Document Metadata — ask questions about the metadata fields
• Document Digest — ask questions about the document index and content
• Draft Answer — get help drafting paragraph responses and affirmative defenses
• PDF viewer Citations panel (Beta) — ask questions about extracted citations on pleadings
• FAQ — get answers about platform features
• Release Notes — ask questions about recent updates

The sidebar supports multi-turn conversations — ask follow-up questions and the AI maintains context from earlier in the chat. Close the sidebar at any time to return to the page.

On most Ask AI pages and all AI Search document list pages (see How does the AI Search feature work on document list pages? above), conversations are saved automatically as you chat. Saved chats are available on:

• Case Notes, Contentions, Case Timeline (List view and Table view), Document Summary, Consolidated Medical Record Summary, Summary Catalog, Document Metadata, Draft Answer, Timesheet Builder, FAQ, and Release Notes
• AI Search on Transcripts, Sessions, Pleadings, Exhibits, Medical Records, Key Documents, and Discovery Documents

The sidebar header includes three controls for managing saved chats:

• New chat (plus icon) — start a fresh conversation without deleting your saved chats
• Chat history (clock icon) — open a dropdown of your saved chats, labeled with the first question and a timestamp; click a chat to resume it
• Delete (X icon on each row in the history menu) — permanently remove a saved chat from your history

Closing the sidebar does not delete saved chats on those pages — reopen Ask AI or AI Search and use Chat history to pick up where you left off. The only Ask AI sidebars that do not save conversations are Case Timeline filter Ask AI (Chart view and List view filter sidebar) and Document Digest — those chats clear when you close the sidebar.