Search a patent and the results come back split into two panels: hits inside the claims, hits everywhere else. The distinction matters — invalidity work is looking for support outside the claims, infringement work is looking for it inside the claims, and the platform sorts results into both buckets so you can read each set separately.

• Per-patent search history that persists encrypted to your account. When you return to a matter months later, the searches you ran the last time on each patent are right there — the cluster of search terms attached to each patent is your matter-specific search vocabulary, ready for more work.
• Progressive search with a word-coverage threshold. Tell the platform how many of the words in your search phrase must appear in any matching sequence in the document — so a search for “pulse width modulation control circuit” with the threshold set at five-of-six still surfaces the patent’s “pulse width modulation control” phrase even though the matching passage omits one word.
• Drag-select any phrase in the patent and the search panel populates and runs immediately — no typing, no copy-paste.
• Multi-term search with semicolon separation; build queries incrementally with shift-drag.
• Cross-patent search delegation across browser tabs — open multiple related patents in separate windows, and the search panel in any of them shows tabs for the others.
• Apply highlight categories to results in another patent from inside your current panel — the colors land on the other patent and stay there, so you can mark up an entire family of related patents from one search session.

Patent Lightning ships with a full suite of pre-defined claim chart templates spanning the analytic operations of patent practice — invalidity, infringement, claim construction, post-grant proceedings, ITC, licensing, portfolio analysis, damages, and more. Each template is a column structure designed for that specific operation, not a generic table you have to retrofit. Every cell is fully formattable in real word-processor controls. Export the finished chart as a real Microsoft Word document, ready to drop into a brief or send to a client.

• Real word-processor formatting in every cell — fonts, colors, bold, italic, lists, alignment.
• Insert and delete rows and columns, merge cells, drag columns to reorder, resize columns to the proportions you need.
• Page setup with size, orientation, and margins; multi-page layout with proper pagination.
• Full undo and redo through complete edit history.
• Real .docx export — an actual editable Microsoft Word document, not a bitmap of one.
• Forthcoming: direct integration with the AI task framework, so AI tasks can produce output formatted into any chart template the practitioner picks.

Patent Lightning’s tasks run against the platform’s own engine by default, with results landing directly in your library. But every task can also be run as a downloadable prompt file you take to whatever AI you prefer — a public chatbot, your firm’s licensed model, an internal corporate AI bound by your data-processing agreement.

The downloaded prompt is the same versioned, layered, citation-strict instruction the platform’s own engine uses, with explicit rules forcing the responding AI to preserve Patent Lightning’s positional citation format. Paste the AI’s response back, click submit, and the response flows through the same pipeline that produces the artifact in your library — citations resolved, output formatted, folder routing handled.

• Your firm’s AI relationship stays your firm’s AI relationship.
• Your work stays under your governance regardless of which AI vendor processes it.
• The platform’s citation discipline carries through to outputs from any AI you bring.
• You are not married to one model, ever.

Type any patent number in any format — granted patent, published application, reissue — and Patent Lightning brings the patent into your library. Paste a list of fifty numbers from a brief and they all come in together. Drop an office action, a brief, a memo, or any other document containing patent numbers into the library, and the platform extracts every valid number with surrounding context, presents them for review with a checkbox per patent, and brings everyone you confirm.

Patent Lightning maintains its own in-house library of the official text and XML files for every U.S. patent and published application from 1976 to the present, refreshed weekly. Text and XML files are not as straightforward to obtain from USPTO as PDFs are — having a curated, indexed, weekly-refreshed in-house library back to 1976 is real infrastructure, and it is what powers the platform’s search index, citation system, claim parsing, and AI surface. The visual PDF document the practitioner reads is fetched live from USPTO.

• In-house text and XML library for every U.S. patent and published application from 1976 to the present, refreshed weekly. The structured data layer that powers search, citation, and AI.
• Live USPTO PDF retrieval for the visual document — the patent the practitioner reads, marks up, and exports.
• Document-driven bulk import from PDF, Word, plain text, and XML — the platform extracts every patent number with 20-word surrounding context, presents them for your review, and brings them all in.
• Full ingestion pipeline for every patent: text aligned to image position, claims numbered, dual-column layout reconstructed in reading order, math and chemistry rendered correctly, every word indexed for citation.
• Forthcoming: an in-house PDF library, eliminating the live USPTO dependency for the visual document as well.

Office actions, prior art memos, draft applications, file histories, third-party research reports, your own working notes — they belong in your patent library too, not scattered across your file system. Upload PDFs, Microsoft Word documents, plain text, Markdown, or rich-text files. They sit in the same folders as your patents, with the same encryption protecting them, available to the same AI tasks you can run against any patent.

• Supported formats: PDF, .docx, .txt, .md, .rtf.
• Same library, same folders, same AI surface as patents.
• Encrypted with your master key the moment they leave your browser.

Some documents do not start as files — they start as a list of patent numbers from a phone call, a draft of new claim language, a memo you are dictating to yourself, a record of an interview, a working summary you build as the matter develops. Patent Lightning gives you a full word processor right in the platform, and what you create saves into your library as a first-class document available to AI tasks the same way an uploaded document is.

• Full ribbon and formatting toolbar — fonts, sizes, colors, alignment, lists, indentation, headings, links, tables, super- and subscript, undo and redo through complete edit history.
• Page setup, page navigation, page indicator with go-to-page, and dual-mode rendering (continuous or paginated).
• Save as Word, save as PDF, save as Patent Lightning document — the editor produces output in the format you need.
• Bidirectional AI integration: editor documents can be selected as targets or references for AI tasks; AI tasks can write outputs that you open in the editor for further work.
• Your work product gets the same place in the library as the documents you receive.

Folders, sub-folders, sub-sub-folders, as deep as you need them. Drag patents and folders between locations; cut, copy, and paste with a real clipboard that knows what it is holding. Bulk-select with checkboxes. The platform does not impose a structure on your matters — you build the one that fits how you actually work.

• Drag-and-drop with shift-to-move; floating tooltip shows copy or move.
• Real clipboard for patents, charts, notes, folders, and mixed selections.
• Bulk operations on multiple items at once.
• Soft-delete to a Deleted Items trash with selective or bulk restore.
• Automatic purge after thirty days; manual purge anytime.

Everything you produce in Patent Lightning can leave Patent Lightning in the formats your firm actually uses. Annotated patents, claim charts, editor documents, AI outputs — all of them export to real Microsoft Word and to real PDF, not to bitmap images of either, ready to drop into a brief, file with a court, send to a client, or attach to an email.

And anything you save can come back. Save any document from the platform to your computer as a self-loading web page; double-click it later and it reopens in Patent Lightning with all of your annotations, highlights, and notes restored. Email the file to a colleague who has Patent Lightning and they see your work the way you saw it.

• Save as Microsoft Word — real .docx output for patents with annotations, claim charts, editor documents, and AI outputs. Editable in Word; not a bitmap.
• Save as PDF — real .pdf output for any document type, with annotations rendered onto the pages and notes arranged in margin boxes you control.
• Save as Patent Lightning document — self-loading HTML that reopens in the application with one click, all annotations and notes restored.
• Full-folder backup as a single zip archive — every subfolder, every patent, every annotation, every chart, every note, decrypted into a portable archive.
• Restore from backup back into your cloud library at any time.
• For matter handoffs, end-of-engagement records, regulatory holds, court filings, client deliverables, or your own peace of mind — your work is yours, your firm’s formats are your firm’s formats, and the platform never gets in the way of either.

Share specific folders with specific people — team members, co-counsel, expert witnesses, or anyone else who needs access to a defined slice of your library. They see what you give them access to and nothing else.

• Per-folder member management with permission levels.
• Concurrent edit handling so two people working on the same matter do not step on each other.
• Per-shared-folder encryption keys — what you share is readable only by the people you have shared it with.
• Patent Lightning itself cannot read shared content any more than it can read your private content.

Patent Lightning is built to be operable by every member of your team, regardless of their accessibility needs. The keyboard commands are not an afterthought layer over a mouse-first interface — they are a parallel command surface designed deliberately, with screen-reader announcements at every step.

• Built-in text-to-speech through the browser engine, with full compatibility for the commercial screen readers your firm already licenses; preferences shared across every page.
• Color-assist mode that adapts the six markup categories for colorblind practitioners.
• Complete parallel keyboard command surface — every menu, every popup, every annotation, every search result, every grid cell in every chart is operable from the keyboard alone.
• Dedicated keyboard modes for creating highlights and cycling through existing ones, with the markup palette navigable as a 6×3 grid via arrow keys.
• Screen-reader announcements at every interaction, with ARIA roles and labels throughout.

Sign in once and every patent, every chart, every editor window you have open recognizes you. Configure your session to whatever fits your workday. Set the platform to lock automatically when you step away from your computer, so client work is protected when you walk to a meeting.

• Configurable session duration from a fifteen-minute session for quick lookups to an eight-hour session for a full day’s work.
• Automatic lock-after-idle at the period you specify, to protect client work when you step away.
• Auto-save across the patent viewer, claim chart builder, and editor, with a visible countdown so you always know when your last save happened.
• Cross-tab session coordination — sign in once, every tab recognizes you.
• File-open conflict detection — when you try to open a file already open in another tab, the platform tells you before you overwrite your own work.