Building the Construction Price Graph
Omnicost is building a graph of catalog items, supplier SKUs, decompositions, regions, and price observations for construction intelligence.
Construction cost data becomes more valuable when it is connected. A price observation is useful. A price observation linked to a supplier SKU, a canonical catalog item, a region, a unit, a decomposition, and a budget line is much more powerful.
Omnicost is building that connected layer as a construction price graph. The graph starts with raw observations from crawlers, catalog imports, BC3 files, supplier pages, and project budgets. It then maps equivalent items together so teams can see one canonical cost concept instead of dozens of disconnected names.
The graph helps answer practical questions. What is the current median price for this item? Which suppliers have moved recently? Which budget rows depend on this category? Which decompositions use this material? Where do we lack regional coverage?
This structure also makes agents more reliable. An agent should not invent prices from text alone. It should retrieve related items, inspect units, compare observations, and explain whether the answer comes from live data, imported catalogs, or a fallback assumption.
For procurement, the graph supports benchmarking. For estimating, it supports faster line-item pricing. For product teams, it creates an API and MCP layer that other tools can query.
The long-term opportunity is a network effect in construction cost intelligence. Every new source improves coverage. Every mapped item improves future matching. Every budget creates feedback that can sharpen the next estimate.