Procore AI

AI layer and agents inside the Procore construction management platform. Vendor says agents execute workflows in Procore through over 150 built-in actions that create or update records such as RFIs, change events, daily logs, and action plans; agents can be triggered by project events or run on a recurring schedule; answers carry numbered citations to source files; permissions follow existing Procore roles; and agents never take action without sign-off. Procore describes a proprietary reasoning model, Magpie, trained on industry data, and says customer data is never used to train models like ChatGPT, Claude, or Gemini. Starter Pack agents listed: Deep Search, RFI (drafts RFIs from project context), Submittal Review (cross-checks submittals against the spec book), Daily Log, and Contract Review (flags scope gaps or risk). Plans page lists large language models from Anthropic and Google on Starter, and Anthropic, Google, and OpenAI on Pro and Enterprise. Customer figures on its AI page (e.g. Haskell cutting submittal review cycles from one week to 10 minutes) are vendor claims, not verified.

Procore customers (GCs, owners, specialty contractors).

Not publicly listed. Not charged per seat. Digital Coworker Starter Pack: 6-month term, flat-rate pricing, up to 3 projects, 5 agents. Digital Coworker Pro (20 agents, video analysis, scheduled automations) and Enterprise (20+ agents, Agent Studio for custom agents, deployment specialist support, BI connection): 12-month terms with credit usage-based pricing; Procore says credits are consumed at data ingest and when agents perform tasks. No prices shown. (Re-checked Oct 15, 2026.)

Prices change; check the vendor page.

Facts last checked: Oct 15, 2026

Not yet tested. No scores exist for this tool.