Overview
If your company already has a proposal layout you like — whether from ArcSite or another tool — you can upload it as a PDF and ArcSite's AI will convert it into a block-based template. The AI analyzes the structure and content of your PDF and maps it to the appropriate block types, giving you a starting point you can refine in the editor rather than building from scratch.
Availability
Supported Platforms: Web (UserSite)
Subscription Requirement: Estimate
Who Is This For?
Admins or template managers who have an existing proposal document they want to bring into ArcSite's block-based template system.
Feature Guide
What It Does
When you upload a PDF, ArcSite sends it to an AI process that reads the layout, identifies sections, headings, content blocks, and form fields, and generates a corresponding block-based template. The template opens in the editor automatically when processing is complete, ready for review and refinement.
In addition to converting the visual layout, the AI also analyzes any form fields it identifies in the PDF and attempts to match them against your organization's existing Field Data questions. Where a match is found, the Field Data question is automatically bound to the appropriate block — eliminating the manual step of finding and linking questions by hand after import.
Where to Find It
Navigate to Settings > Templates > Proposal Templates, click Create Block Template, and select Create with AI.
How to Use It
Step 1: Upload your PDF
Select a PDF file from your device. The file must be 5 pages or fewer. Larger documents are not supported for AI conversion.
Step 2: Wait for processing
AI conversion runs as a background task. Processing time will vary depending on the complexity of your document. You do not need to stay on the page while it processes.
Step 3: Review the generated template
When processing is complete, the template opens in the block editor. Review each section to confirm the AI has interpreted your layout correctly. Pay particular attention to:
Headings and section breaks
Any blocks where content may have been combined or split unexpectedly
Image placeholders — the AI will insert placeholder blocks where images were detected, which you will need to replace with the appropriate image blocks manually
Step 4: Review Field Data matches
If your PDF contained form fields, check the blocks where Field Data questions have been automatically bound. Confirm each match is correct and adjust any bindings that don't align with your intent.
Note: Field Data auto-matching works against your organization's existing Field Data question bank. If a relevant question doesn't exist yet, you'll need to create it in Field Data settings and bind it manually.
To learn how to bulk import and create field data please refer to this article: Bulk Import Field Data or Set Up Field Data
Step 5: Finish editing and publish
Make any remaining adjustments in the editor, then publish the template when it's ready.
FAQs & Troubleshooting
Q: What is the PDF page limit? PDFs must be 5 pages or fewer. If your existing proposal is longer, consider splitting it or uploading the most structurally representative pages.
Q: Will images from my PDF carry over? Images are not directly imported. The AI inserts placeholder blocks where images were detected. You'll need to replace these with the appropriate image blocks in the editor.
Q: What if the AI gets the layout wrong? The generated template is fully editable. Think of the AI output as a starting draft — you can adjust, add, remove, and reconfigure any block after import.
Q: What if none of my Field Data questions get matched? The AI matches against your existing Field Data bank. If your organization hasn't set up Field Data questions yet, no matches will be made. Set up your Field Data questions first, then re-import if needed.
Q: Can I use this to convert a legacy ArcSite template? Yes. Export an existing proposal generated from a legacy template as a PDF, then upload that PDF through this flow. See Using the Block-Based Template Editor for more context on the relationship between legacy and block-based templates.