Trunk Tools Launches Cortex: An AI Platform for Construction Documents

What Cortex Actually Does

Cortex is an AI platform designed specifically for construction workflows. It reads across the full document stack—drawings, specifications, RFIs, submittals, schedules, and change orders—and connects them in a single interface so teams can trace project impacts without hunting through folders.

The platform powers several AI workflow agents, each targeting a specific pain point: drawing revision analysis, RFI management, bid analysis, and submittal reviews. Think of them as specialized assistants that actually know what a submittal is, rather than a general-purpose chatbot pretending to.

The Detail That Matters Most

One capability worth flagging: Cortex is built to detect drawing revisions that fall outside traditional revision clouds. That’s the kind of change that slips through manual review, gets missed, and quietly becomes a rework problem three months later.

It’s a narrow feature, but it’s exactly the kind of thing that separates a construction-native tool from a generic document AI with a hard hat slapped on it.

Four Years in the Making

Trunk Tools says Cortex was developed over four years using data from real commercial construction projects. That’s a meaningful signal. Construction has its own vocabulary, its own document conventions, and its own failure modes—none of which a general-purpose model learns from a Wikipedia crawl.

The specificity of the training data is arguably the product’s strongest differentiator.

Who This Is For

Cortex is aimed at construction teams dealing with high document volume—general contractors, project managers, and anyone whose job involves reviewing bids, tracking RFIs, or catching the drawing change that everyone else missed.

If your current workflow involves downloading PDFs, cross-referencing specs manually, and hoping nothing slipped through, this is worth a close look.

The Takeaway

Cortex isn’t trying to be an AI for everything. It’s trying to be the AI that actually understands a construction project—and that focus is exactly what makes it credible. The real question isn’t whether it works in demos. It’s whether it holds up when a 400-sheet drawing set lands on a Monday morning.