GHG Inventory: Auditable and in Real Time.

Carbon Intelligence is the greenhouse gas inventory agent in CarbonOS, Carbonova’s platform of AI agents. It calculates Scope 1, 2 and 3 emissions with the GHG Protocol method from data the company already holds in ERP, SCADA, e-invoices and IoT sensors, flags anomalies automatically and records every calculation in an audit trail.

It replaces the GHG inventory assembled by hand in spreadsheets, which ties up the team at every close, depends on a few people and is hard to defend under external verification. It serves sustainability, energy, finance and compliance teams in industries such as cement, steel, mining, chemicals, oil and gas and food, which must give regulators, customers, lenders and investors the same number.

Activity data arrives through Signal Ingest: fuel and energy from SCADA and meters, purchases and invoices from the ERP (such as TOTVS, Senior, SAP or Oracle), freight from TMS/WMS and IoT readings. The agent applies the right emission factor to each data point, with its version recorded, totals the results by scope and category and flags unusual values for review before the inventory is closed.

Yes. It produces the Scope 1, 2 and 3 inventory used in the climate section of a CSRD report, under the revised ESRS that apply from financial year 2027. For Scope 3, mind the value-chain cap: from 2027, a company in scope may not ask suppliers with up to 1,000 employees for more information than the EU voluntary standard provides, so supplier requests should stay within it.

It receives normalised data from Signal Ingest and uses Asset Graph to know which plant, line and piece of equipment each emission belongs to. Proof Engine keeps the lineage and versioned factors behind every number; Compliance Hub turns the inventory into reports; Supply Chain brings supplier data into Scope 3; and Decarbonization Planner uses the same dataset for scenarios and SBTi-aligned targets.

With a spreadsheet, every close starts almost from scratch: someone copies data from several systems, updates factors by hand and the history of changes gets lost. In Carbon Intelligence, data comes straight from the source systems, every factor has a version, every adjustment is logged with its author and date, and anomalies surface during the year, not only at the close. Any number can be recalculated and explained to the verifier.