Persefoni Improves Data Accuracy With New AI Functionality For Carbon Management

Persefoni Improves Data Accuracy With New AI Functionality For Carbon Management

2023 may be the ultimate year of AI breakthroughs. Amidst advances from technology giants like Meta and Microsoft, carbon management software vendors too have begun to incorporate elements of AI into their solutions to automate, enhance and streamline data management and analysis processes. For example, US-based climate software vendor Persefoni has recently launched Persefoni AI, a suite of embedded AI capabilities designed to increase data accuracy and ease data management logistics. Last week, they announced their latest AI-based innovations.

Persefoni had already been using AI within its platform to map data types more accurately. The new functionalities will further aid users by automatically identifying data errors, assessing potential data anomalies/outliers and allocating appropriate emission factor recommendations. Error detection automates standardized data formatting, reducing the time needed for data management. This ensures that data provided from multiple sources is compatible, with consistent unit measurements, allowing more efficient data flow and analytical processes. The anomaly detection function can identify outliers in datasets without the need for manual inspection. This AI function uses previously ingested data to identify potentially incorrect data from regular input sources. Currently both functions rely on the user to review the identified errors and anomalies and select one of the range of suggested corrections. However, there are plans to further expedite this process by automating data correction based on previous inputs.

Emissions factor allocation is increasingly an area of differentiation for carbon management software solutions; generating accurate secondary data requires access to a substantial quantity of granular emission factors to correspond with specific activity data. Vendors usually achieve this through integration with vast emission factor databases, such as those belonging to Ecoinvent, the IPCC, the UK’s DEFRA and the EPA in the US. However, without an automated mechanism to assign the most appropriate emissions factor, this creates a trade-off for the user between accuracy and logistical burden. Persefoni saves users from sifting through thousands of emission factors by training the platform to provide a list of the most suitable factors based on the data type and business descriptions. Recommended factors are ranked by suitability and given a similarity score.

These types of automated functions are likely to become increasingly useful as users are required to collect a larger volume of data from a substantial number of sources across a firm’s value chain. For more information regarding best practices for accessing solution functionality and RFP creation, see Verdantix Best Practices: Creating An RFP For Enterprise Carbon Management Software.