What’s New in Parse: Automated Construction Takeoff, CSI Filtering, and Smarter Document Outputs
This release expands how your team interacts with documents at the project level, adds branded output capabilities for client-ready deliverables, and ships a set of performance improvements that make the platform faster and more reliable across the board.
Project-Level AI Document Chat: Ask Questions Across Your Entire Plan Set
Previously, AI Document Chat was scoped to a single document at a time. This release changes that.
What changed:
AI Document Chat now operates at the project level, meaning you can ask questions that pull context from every document in your project simultaneously.
Why it matters:
Construction projects do not live in one file. Specs, drawings, addenda, and sub quotes are spread across a full plan set. Project-level AI chat means an estimator can ask “what are the door hardware requirements across all drawings” or “are there any conflicts between the mechanical spec and the electrical drawings” and get an answer that draws from the full document set, not just whichever PDF happens to be open. This is a significant step forward for pre-construction document automation on complex, multi-document bids.
Branded Export: Send Parse Outputs With Your Firm’s Name on Them
What changed:
Parse exports now support branding customization. You can add your company logo, set a custom header, apply name substitutions to relabel default field names to match your firm’s terminology, and adjust surrounding text to match your internal or client-facing format.
Why it matters:
Whether you are dropping a scoped quantity list into a bid package, sharing results with a sub, or presenting takeoff data to an owner, the export now reflects your firm rather than the tool that produced it. For pre-construction teams that use Parse outputs as deliverables, this removes the reformatting step that used to happen after export.
Performance and Reliability Improvements
Error Resolution
What changed:
Continuous errors that were appearing in production have been resolved.
Why it matters:
These errors were creating intermittent friction across the platform. Fixing them makes every interaction slightly faster and removes a source of unexpected behavior that some users were seeing without a clear explanation.
Document Preparation Optimization
What changed:
Adjustments to the preparation service improve how documents are processed before they are ready for takeoff and AI interaction.
Why it matters:
Faster, more consistent preparation means your documents move through the pipeline more reliably, particularly on larger or more complex file sets.
These updates are live now!
Project-level AI chat, branded exports, and the performance improvements in this release are all live in the sandbox. Upload a multi-document project, run a cross-document question, and export the results with your logo on them. Tell us what you find.
Send us your worst document
