I, Librarian is an online service designed to streamline the management, organization, and enrichment of your professional document collections, including PDF papers, research articles, and essential office documents.
It functions as a comprehensive productivity hub, offering advanced features specifically tailored for:
This documentation is structured to help you quickly understand the platform, master key workflows, and become highly efficient using I, Librarian's organizational, collaborative, and search capabilities.
I, Librarian features a highly consistent user interface (UI) structure across all views, minimizing the learning curve and making workflows predictable. (See the accompanying diagram below for visual reference.)
| UI Component | Location | Function |
|---|---|---|
| Navigation panel | Left side | Provides immediate access to major library sections, groups, and settings. |
| Main content area | Right side (largest section) | Displays documents, search results, metadata, and active views. |
| Contextual toolbar | Bottom of the main content area | A dynamic toolbar that changes based on your current view, providing essential quick actions. |
The toolbar provides quick access to actions relevant to the content you are viewing, including:
I, Librarian features a secure and streamlined system for managing user access, authentication, and account creation.
When I, Librarian has no existing accounts, you will be greeted by a Create account page. The registration process requires several mandatory pieces of information to ensure security and uniqueness. The system also includes a robust password strength meter that warns you if a password has been compromised in public data breaches.
It is recommended that once the first account is created, user registration be disabled in Administrator > Global settings. New user accounts can then be created on the Administrator > User management page instead.
If you have forgotten your password, the system requires an email verification step before allowing you to set a new one.
I, Librarian organizes your resources into four distinct collection types. Understanding the purpose and limitations of each collection helps you manage your workflow and collaboration efforts efficiently.
Tip: For quick navigation while reading, the following keyboard shortcuts can be used to move between pages: W or A navigates to the previous page, S or D navigates to the next page.
Library is the permanent home for all documents users have imported into I, Librarian. Here you can search, filter and browse all these items. This whole collection is shared among all users.
Clipboard is your temporary staging area, a transient holding space for documents you have recently selected or are preparing to act upon (e.g., tagging, moving, or bulk exporting). This collection is private, visible only to the account that created it, and persists across devices and sessions until manually cleared. To add an item to the Clipboard, click the dedicated Clipboard checkbox located under the item's title in any list view.
Projects are task-focused workspaces, designed to group documents relevant to a specific research goal, paper, thesis, or collaborative effort. Projects can be shared with specific collaborators.
The Projects page is your central hub for managing the project collections. Here you can find:
Each project is accompanied by contextual actions:
You can create a new project in the Create new project card:
Adding items to a Project allows you to use several specialized tools for collaboration, communication, and document synthesis:
| Project chat | AI-powered Q&A derived from the project documents (currently only titles and abstracts are used). Chat threads are shared among all users. |
| Project summary | A feature that utilizes AI to analyze the content of all documents (currently only titles and abstracts are used) within the project and generate an overarching summary or abstract. Summaries are shared among all users. |
| Project notes | A dedicated section for rich-text note-taking, drafting outlines, saving research thoughts, related to the project scope. Project notes are shared among all users. |
| Note compilation | This tool aggregates all notes from individual documents into a single list, ideal for review and writing. |
| Discussion | A message board for longer-form conversations, decision tracking, and discussing broader project strategies among collaborators. |
Catalog provides a comprehensive view of every document. This collection is necessary because the default collection listings (Library, Projects, etc.) are optimized for speed and only display the top 10,000 items. The Catalog allows you to find any item that would otherwise exceed this item limit.
I, Librarian allows you to expand your document collection by searching and importing references directly from major academic, scientific, and technical databases worldwide. You can initiate a search across any of the following repositories directly from the I, Librarian interface:
| arXiv | A large repository of over 2 million electronic preprints approved for posting after moderation, focusing on physics, mathematics, computer science, and related fields. |
| Crossref | An official Digital Object Identifier (DOI) Registration Agency containing metadata for over 100 million articles from academic journals. |
| IEEE Xplore | A comprehensive research database containing over 6 million documents on computer science, electrical engineering, and electronics. |
| NASA ADS | The Astrophysics Data System, an online database of over 15 million astronomy and physics papers from both peer-reviewed and non-peer-reviewed sources. |
| Pubmed | Comprises more than 36 million citations for biomedical literature from MEDLINE, life science journals, and online books. |
| PMC | A free digital repository that archives over 8 million publicly accessible full-text biomedical and life sciences articles. |
| Patents | Contains data on more than 150 million patent documents from the European Patent Office (EPO) and major international patent offices. |
| ScienceDirect | Elsevier's database containing over 20 million pieces of content from academic journals and e-books. |
| Scopus | Elsevier's citation database containing over 94 million records covering life, health, physical, and social sciences. |
This method bypasses traditional keyword searching and directly fetches the metadata for the specified documents using unique identifiers (UIDs) that you already possess.
The "Import PDFs" function is used to upload published PDFs containing DOIs. I, Librarian automatically
attempts to extract the DOI from the uploaded file and match it against external databases. In addition,
if the PDF filename is a URL-encoded DOI or a PubMed ID (PMID), I, Librarian will directly use this UID
to look up and enhance the document record. Example: 10.1002%2Fadvs.202412850.pdf, or
PMID39887888.pdf
You have two methods for supplying your PDF files:
https://) into
the URL field if your file is hosted online.
After uploading, I, Librarian processes the file to find identifiers (like the DOI) and then searches external databases to enrich the record. You can prioritize which database is used for metadata retrieval: IEEE, NASA ADS, Pubmed, Pubmed Central, and Crossref.
Before clicking Save, you can choose to assign the imported PDF(s) to one or more existing Projects and Clipboard. Also, you can apply specific Tags to the new PDF(s) for enhanced organization and filtering.
It is highly recommended to leave the don't save duplicates box checked. I, Librarian will compare the uploaded file's metadata against your existing Library content and skip the import if an exact match is found, preventing clutter.
The "Import metadata" form is used to transfer references from other citation managers or databases using standard metadata formats: RIS, BibTeX, and Endnote XML. You have three options for importing your metadata:
Before clicking Save, you can choose to assign the imported references to one or more existing Projects and Clipboard. Also, you can apply specific Tags for enhanced organization and filtering.
It is highly recommended to leave the don't save duplicates box checked. I, Librarian will compare the uploaded file's metadata against your existing Library content and skip the import if an exact match is found, preventing clutter.
If you are importing a large collection of references that also have associated PDF files, I, Librarian
can automatically upload those PDFs, provided the metadata file includes a relative file path and the PDFs
are located in your I, Librarian data/import directory.
The Manual import feature is used for adding files that are not published articles (e.g. proprietary internal documents, drafts, or image files) or for files where automated metadata matching is not necessary or possible. This method allows you to define all necessary details and file uploads in one view.
Before clicking Save, you can choose to assign the imported document to one or more existing Projects and Clipboard. Also, you can apply specific Tags for enhanced organization and filtering.
Tip: You can open the search modal window in any list view by pressing the F key.
The Quick search modal provides immediate access to your Library content using simplified search parameters. It is designed for fast retrieval of the most relevant documents based on keywords or specific identifiers.
The Quick Search interface features predictive text and live result fetching, allowing you to find documents quickly without needing to execute a full search query. As you begin typing your query, the system provides immediate live suggestions:
Check the save this search for later box to save your current search query. Saved searches are accessible from the Previous searches link, allowing you to quickly rerun complex or frequently used queries.
The Advanced search gives you precise control over your queries, allowing you to combine multiple criteria (e.g., author, year, and specific keywords) using Boolean logic to narrow down results.
Advanced Search uses multiple query lines, each consisting of three parts: boolean operator, content type, and search query. To add or remove these lines, click the blue plus sign and grey minus sign, respectively.
You can define the order in which the search results will be displayed:
Check the save this search for later box to save your current search query. Saved searches are accessible from the Previous searches link, allowing you to quickly rerun complex or frequently used queries.
The Export modal allows you to extract documents (those in the Clipboard, a Project, or a Search result set) in various metadata formats or as a complete offline package.
Select one of the following formats based on your destination application or need:
| Format | Description | Primary Use |
|---|---|---|
| BIBTEX | The standard text-based format widely used with LaTeX environments. | Integrating references into LaTeX documents. |
| BIBTEX + abstracts | Includes the standard BibTeX data plus the abstract text for each document. | Full reference sharing with colleagues. |
| Endnote | A proprietary format for Thomson Reuters EndNote citation software. | Migrating data to EndNote. |
| RIS | A standardized tag format used by most major reference management programs. | Migrating or sharing data with non-LaTeX citation managers. |
Select citation style to export formatted citations based on a specific Citation Style Language (CSL) format. Use the Search field to find the specific journal style you wish to use for formatting.
Select offline app to export PDFs in a zip file containing the selected PDFs, and metadata into a single compressed file designed for use offline.
PDFs can be also downloaded by clicking on a button with a download icon in all PDF button groups.
When you click on a document title in a collection list, the view shifts to the detailed Item View. The navigation panel updates to provide specialized tools for interacting with that specific document, its metadata, and its associated files.
I, Librarian allows several ways to write notes about an item. They can be accessed in the navigation under:
The Tags page allows you to organize your document by associating it with relevant keywords and concepts. This view is divided into three sections: manual tag management, AI-suggested keywords, and a comprehensive list of all existing tags in your Library.
To add new (non-existent) tags use the Add New Item Tags form. Type one tag per line into the text box. If the tag already exists in your Library, it will be automatically linked; otherwise, a new tag will be created.
I, Librarian automatically analyzes the content (title, abstract) of the document to suggest highly relevant keywords in the AI Keywords list. Click the Use button next to any suggested keyword to automatically add it to the Add new item tags field.
The Item Tags section provides a view of the entire vocabulary of tags currently in use across your Library and allows for efficient application of existing tags. Recommended tags are highlighted in gray.
The Supplements section is dedicated to linking ancillary files and data (such as raw data sheets, high-resolution figures, presentations) to the main document record.
You can upload supplementary files using two methods located in the left panel. Use the dashed box ("Click to select | Drop files here") to upload files directly from your computer. Alternatively, paste the direct link into the URL field to fetch a file from an external source. Check the graphical abstract box, if the file you are uploading is intended to be used as the graphical abstract.
Existing supplementary files are categorized and listed in the right panel, allowing for easy review, download, and modification. For every listed supplementary file, the following controls are available:
I, Librarian offers advanced Artificial Intelligence tools to accelerate your research by analyzing the content of individual PDFs, generating summaries, and identifying related materials. These tools are accessible from the detailed Item View navigation pane.
The PDF chat function turns your document into an interactive knowledge base, allowing you to converse with the document's content. Ask natural language questions about the PDF (e.g., "What were the main findings?" or "What methodology did they use?"). The AI analyzes the full text and provides answers sourced directly from the document. Located in the footer, you can select the AI model and adjust its behavior. A dedicated input field Ask a follow-up question allows for continuous conversation and refining previous questions.
The PDF summary tool leverages AI to generate an abstract or condensed version of the document, saving significant time in determining relevance and key findings. This feature is particularly useful when the imported document lacks a formal abstract (older papers, technical reports), or when the existing summary is brief, generic, or of low quality (common with patent records or some non-journal publications).
The Similar articles tool uses AI to deeply analyze the content of the current paper and generate highly specific search queries that lead to related documents within the broader scientific literature. After running the Generate search queries function, the Similar articles page displays a list of pre-optimized search strings designed to find highly relevant literature. Queries are complex, using nested parentheses and Boolean operators (AND, OR) to combine highly specific terms. Once you have selected a radio button, the search string is ready to be executed against both internal and external resources using the Search selected query links.
The Manage PDF view provides control over the physical file associated with the document record, allowing you to update the PDF and manage its indexed text.
If you have a newer or better-quality version of the associated PDF (e.g., a final copy replacing a pre-print), you can replace the file without altering the document's metadata record.
The right panel shows the currently indexed text that I, Librarian uses for full-text searching. If the extracted text is empty, or poor quality (indicating the original PDF was a scanned image without proper text layers), clicking OCR will trigger a process to analyze the image of the PDF and generate new text.
I, Librarian features a powerful, custom PDF viewer built directly into the application, designed for detailed reading, annotation, and efficient document navigation. The viewer is fully integrated, ensuring seamless transitions between PDF reading and other native tools.
Left panel can display thumbnail previews of all pages, bookmarks, user notes, and search results. Main display renders the PDF pages. The top toolbar provides navigation and annotation functionality:
Tip: Pressing the H key will immediately hide all surrounding menus and toolbars, providing a clean, uninterrupted view of the PDF content. This is especially effective when combined with the browser's full-screen mode (usually toggled by pressing F11).
Tip: For quick navigation while reading, the following keyboard shortcuts can be used to move between pages: W or A navigates to the previous page, S or D navigates to the next page.
The Omnitool is a powerful feature that enables you to perform bulk actions on all items (not just currently visible) in the active list view (e.g., all documents in a search result, a Project, or the entire Library view).
The Omnitool can be used to quickly assign a large import set or a filtered group of papers to a specific Project, mass tagging or untagging documents after a comprehensive search, using the Clipboard actions to quickly create sub-lists for further processing or batch export, etc...
The Duplicates section provides tools to scan your entire Library and identify potentially redundant document records. This is helpful for maintaining a clean, efficient, and accurate research collection. Duplicate detection uses three distinct methods:
When you click Find duplicates under any of the three categories, the system presents the detected duplicate groups for resolution. For each group of duplicates, the system requires you to select one item to keep. I, Librarian will try to merge all metadata, notes from the deleted files into the preserved item.
The Single Sign-On (SSO) settings allow I, Librarian to integrate with your organization's centralized identity provider (IdP). This enables users to access I, Librarian using their existing corporate or institutional credentials.
SSO login toggles control which authentication methods are active:
Warning: It is critical to ensure that SAML or OIDC login is fully functional before switching Local login off. Disabling local login prematurely may lock out all users, including administrators.
The SAML tab provides fields for integrating with a SAML Identity Provider (IdP):
| Service provider (I, Librarian) | Entity ID, ACS link | Information about I, Librarian that must be registered in your organization’s IdP. |
| Identity provider (SAML service) | Entity ID, IdP URL, TLS certificate | Details and security credentials from your IdP that I, Librarian requires to trust the authentication. |
| Map user attributes | Email, Username, First name, Last name | Defines which SAML attribute fields map to the corresponding I, Librarian user profile fields, ensuring accurate account creation and linking. |
| Map user groups | Admin group, User group, Guest group | If provided, incoming user permissions will be automatically set based on the group claims received in the SAML payload (e.g., users in the "Admin group" claim receive admin privileges in I, Librarian). |
The OIDC tab provides fields for integrating with an OpenID Connect Identity Provider:
| OpenID Provider (SSO service) | Issuer URL, Authorization URL, Token URL, Keys URL, User info URL | Endpoints required by I, Librarian to interact with the OIDC provider for user verification and token exchange. |
| Relying Party (I, Librarian) | Client ID, Client secret, Redirection URL | Credentials and URLs that uniquely identify your I, Librarian service location as a trusted client application within the OIDC provider. |
| Map user groups | Admin group, User group, Guest group | Defines how group claims passed via OIDC will automatically assign permission levels within I, Librarian. |
Tip: If your Identity Provider implements the standard OIDC discovery specification
(.well-known endpoint), simply pasting the Issuer URL may automatically populate the remaining critical
endpoints (Authorization URL, Token URL, etc.), significantly speeding up the configuration process.
I, Librarian offers a simple Application Programming Interface (API) that allows developers and administrators to integrate the platform with external services, scripts, and custom applications. The full API documentation, including all available endpoints and request parameters, can be found directly on the Administrator > API page.
Access to the API is secured using key-based authentication combined with optional IP address restrictions, ensuring only trusted sources can interact with your Library data.