AI Historical Cost Search Agent

How Cassidy automates using AI

Step 1: Ingest historical cost data

The Workflow triggers when new cost data arrives—whether from uploaded estimate workbooks, exported bid tabs, ERP feeds, or subcontractor quote submissions—and parses line items, quantities, UOM, labor/material/equipment costs, and project metadata.

Step 2: Normalize and map to industry standards

Cassidy adjusts historical unit prices to a common pricing date using escalation indices, applies location factors (RSMeans CCI or locally researched costs), harmonizes UOM and scope differences, and maps line items to CSI MasterFormat or UniFormat codes.

Step 3: Enrich with RSMeans benchmarks

The Workflow pulls licensed RSMeans unit prices and assemblies, applies the appropriate City Cost Index for the target location, and stores crew composition and productivity assumptions to explain any variances between internal costs and market references.

Step 4: Detect anomalies and validate

Cassidy flags line items whose normalized costs fall outside expected bands given scope, location, and wage regime—surfacing warnings when scope differs (fire-rated vs standard, premium manufacturers) and highlighting volatile material trends.

Step 5: Serve insights in the estimator's workflow

When an estimator searches a scope, Cassidy returns historical ranges, recent medians, RSMeans benchmarks with CCI factors, and productivity notes—allowing one-click acceptance or adjustment with auto-generated Basis of Estimate documentation.

Step 6: Generate bid artifacts

The Workflow pre-populates line items for common scopes, applies markups and coefficients, produces alternates and add/deduct schedules, and rolls up OH&P, bonds, and contingency—with full audit trails and source citations attached.