Overview
RelayHub lets you attach files directly in chat and have the AI analyze them on the spot. Drag and drop a PDF, spreadsheet, image, or document into the chat input, and the AI automatically extracts content, recognizes structure, and responds with analysis. No separate upload step or external tool required.Supported File Types
- Documents
- Spreadsheets
- Images
- Data Files
| Format | What AI Extracts |
|---|---|
| Full text extraction with layout preservation. Tables are detected and structured. Scanned PDFs are processed with OCR. | |
| Word (.docx) | Text, headings, lists, and tables. Formatting context is preserved so the AI understands document structure. |
| PowerPoint (.pptx) | Slide content including titles, body text, speaker notes, and embedded tables. |
How to Attach Files
Drag and drop
Drag a file from your desktop or file manager directly into the chat input area. A drop zone indicator appears when the file is over the input.
Or click to browse
Click the Attachment icon (paperclip) in the chat input to open a file browser. Select one or more files.
Add your question (optional)
Type a message alongside the attachment to guide the AI’s analysis. For example: “Summarize the key findings in this report” or “What trends do you see in this data?” If you send the file without a message, the AI provides a general summary.
What the AI Does with Your Files
PDF and Word Documents
The AI extracts all text content and understands the document’s structure — headings, sections, tables, and lists. You can ask it to summarize the document, find specific information, compare sections, or rewrite content. For PDFs with tables, the AI uses advanced table extraction to preserve row and column relationships.Excel and CSV Spreadsheets
Spreadsheet files get special treatment. The AI parses the data into structured form and can:- Answer natural language questions about the data (“What was total revenue in Q3?”)
- Generate statistical summaries and highlight patterns
- Create charts and visualizations from the data
- Perform calculations, aggregations, and comparisons across columns
For large spreadsheets (thousands of rows), the AI processes data in batches to manage memory. You can still query the full dataset — the batching is handled automatically behind the scenes.
Images
When you attach an image, the AI uses vision capabilities to understand what the image shows. This includes:- OCR — Extracting text from photos of documents, receipts, signs, or screenshots
- Chart recognition — Reading bar charts, line graphs, pie charts, and other data visualizations, then describing the data they represent
- Diagram analysis — Understanding flowcharts, org charts, architecture diagrams, and other visual structures
- General analysis — Describing what is in the image and answering questions about it
PowerPoint Presentations
The AI reads through all slides and extracts content including titles, bullet points, speaker notes, and embedded visuals. You can ask it to summarize the presentation, extract key takeaways, or critique the argument structure.Inline Data Analysis
When the AI identifies tabular data in your files, it can generate charts and visualizations directly in the chat. These include bar charts, line charts, scatter plots, histograms, and more. The charts render inline in the conversation — no need to download or open a separate tool.File Size and Limits
- Maximum file size per attachment depends on your subscription plan
- Multiple files can be attached in a single message
- Uploaded files are stored securely and are only accessible to participants in the conversation
- Files attached in chat are separate from files uploaded to the File Hub — chat attachments live within the conversation context