Unlock PDF Power: Centralize, Analyze, and Share Documents Using AI!

A PDF report in your email, a reference article in a tab, notes in a Word file, a spreadsheet somewhere in the cloud: the daily document routine of a digital professional often resembles a never-ending treasure hunt. And when it’s time to cross-reference these sources to create a synthesis, the time spent gathering everything sometimes exceeds the time dedicated to the analysis itself.

This is the challenge that PDF Spaces aims to address, a topic we explored in a previous article. The tool allows you to centralize up to 100 sources in a single space, then query them through a conversational AI assistant. Each response is based solely on the provided corpus and directs you to the exact passage in the original document. Here’s how to utilize it further, from creating a space to sharing it with a team.

Learn more about PDF Spaces

How to Create and Populate a PDF Space

Setting up a PDF Space takes just a few seconds. On a desktop, simply select “PDF Spaces” from the Acrobat homepage, then click on “Create a PDF Space.” The same feature is available on the web and mobile (iOS and Android), allowing you to build or supplement a corpus on the go by adding documents directly from a mobile device.

Regarding formats, PDF Spaces supports a wide range. It accepts PDF, Word (DOCX), Excel (XLSX), PowerPoint (PPTX), text (TXT), RTF, and even subtitles (VTT), as well as web links and text copied and pasted directly into the interface. Importing is done by drag-and-drop, from cloud storage (Google Drive, OneDrive, Dropbox, Box), or by simply pasting URLs. On Windows, an option even allows you to create a PDF Space directly from an open Word document with a single click. The search for pages via URL and document upload are automatic and take just a few seconds.

As for languages, PDF Spaces supports English, French, Spanish, Italian, and Portuguese, allowing you to build and utilize corpora in multiple languages within the same space.

A tip: clearly naming each space (by project, client, or theme) facilitates navigation when you accumulate multiple. It is also possible to add a description to each PDF Space to specify its scope. Additionally, Adobe offers a help page to assist with space creation.

How to Query Your Documents with the AI Assistant

Once the sources are loaded, the AI comes into play. The first level is automatic. As soon as the corpus is formed, the assistant generates a global summary, identifies recurring themes, and highlights key points in the form of insight cards. This initial overview allows you to gauge a corpus without opening a single file.

The title of the PDF Space is also generated automatically based on the added sources (it remains fully editable). The assistant also suggests recommended goals to help start exploring the corpus, which proves useful when you are not yet sure which angle to prioritize.

A Conversational Layer

The second level is conversational. The chat interface allows you to ask questions in natural language about all documents. You can request a cross-summary (“What are the common points between these three reports?”), extract specific data (“What is the revenue mentioned in document X?”), or identify contradictions between sources.

For those unsure where to start, the assistant offers suggested questions during and after document processing, as well as prompt ideas for exploring the corpus without having to formulate your own queries. The interaction does not only occur through the chat panel: a prompt bar, integrated directly into the interface, enables quick actions (summarize, analyze, generate) without opening the assistant panel.

Source-Verified Exchanges

Each AI response includes clickable citations. A click takes you directly to the relevant passage in the source document, allowing you to verify information in context effortlessly. This mechanism is central to Adobe’s approach: the assistant does not just respond, it justifies each element presented.

An important point: the AI works exclusively on the provided corpus. It does not perform any web searches or draw from any external knowledge bases. This closed-loop operation ensures that responses remain relevant to the defined scope and that no sensitive data is transmitted outside the space. Adobe also specifies that it does not use uploaded content to train its AI models.

How to Customize the AI Assistant for Your Needs

One of the most useful, and perhaps least known, aspects of PDF Spaces is the customization of the AI assistant. By default, three specialized assistants are offered in addition to the main assistant: Analysis (focused on critical thinking and hypotheses), Training (breaking down complex topics into accessible points), and Entertainment (creative thinking, wit). But the real value lies in the ability to create your own assistants by specifying specific instructions.

An Infinity of Analytical Prisms

In practice, you just need to click the dedicated button in the prompt box, select “Create your own assistant,” and describe in a few lines the desired analytical prism. A marketing manager might ask the assistant to focus on business implications and market data. An HR professional might configure it to respond to employee questions about internal processes, based on company documentation. A consultant might tailor the analysis to the profile of their client or sector.

The effect is immediate, the automatically generated insights refresh to reflect the new prism, and the conversational responses adopt the requested reading level. The same financial data corpus will not be analyzed in the same way whether it is addressed to a CEO or a management controller. The configuration takes a few seconds and can be changed at any time.

Save and Organize Your Findings with Notes

As you explore, the relevant responses from the AI assistant can be saved in Notes, a feature integrated directly into the PDF Space. Each note is editable: you can add your own thoughts, complete an analysis, or sketch out an action plan. These notes are kept in the space and can be copied into the chat to restart a conversation or serve as the basis for a deliverable (report, project, brief). When a PDF Space is shared, the notes are visible to collaborators, making it a collective work tool.

How to Share a PDF Space and Collaborate as a Team

PDF Spaces is not designed for individual use only. Each space can be shared at three access levels:

• Private (only invited individuals),
• Organization (accessible to all members of the same organization),
• Public (open to anyone with the link, no registration required).

Permissions are, of course, configurable. A collaborator can be granted read-only access, the ability to comment, or full editing rights, which include adding files and notes. Recipients do not need a subscription to access a shared space, which facilitates exchanges with external parties (clients, service providers, partners…).

Once invited via the “Share” button, collaborators can review all sources, browse the generated insights, access specialized assistants and notes, and pose their own questions to the AI assistant. However, the chat history remains private: each user interacts with the corpus in their own conversation thread, without seeing the exchanges of others.

For sensitive documents, it is recommended to use private links with named recipients and avoid public links that provide access without authentication.

Discover PDF Spaces