How Carbon Accounting Firms Reduce Document Onboarding Time from Months to Hours

About the client:

The client is a leading carbon accounting software in the US serving many F500 enterprises sustainability team. They quantify businesses' carbon footprint by collecting utility data such as electricity, gas, and water usage. By accurately quantifying GHG emissions, they help businesses identify opportunities to comply with regulatory requirements, reduce their carbon emissions, and set sustainability goals. They are headquartered in London but have expanded their operations to the US.

Challenge:

Calculating GHG emissions requires manually analyzing data related to Scope 1, 2, and 3 emissions. These scopes are the standard frameworks which help categorize emissions into direct and indirect categories to better understand a businesses’ carbon impact. A primary source for collecting this data is utility bills detailing the amount of electricity, gas, water, and waste generated by the business. Utility bills come in various formats that include utility types, meter readings, usage history, usage breakdown, peak demand, efficiency metrics, and waste amounts. The complexity of bills requires contextual understanding of utility relevance to the business and in-depth manual review to accurately extract energy consumption data for calculations.

Key Challenges:

• Data variability: Utility bills are difficult to standardize given their varied formats and terminologies used across different businesses. This can lead to inaccurate GHG emission calculations.

• Scalability Issues: Extensive time is required to accurately process thousands of documents manually.

• Timeliness of Insights: Delayed insights can result in businesses' inability to meet regulatory compliance and hinder effective carbon management strategies.

Solution:

Trellis’s advanced AI helps streamline the process of consolidating unstructured energy use data for carbon management. Trellis turns utility bill information into SQL-compliant format, allowing the company to identify the data schema they are interested in, using natural language. Trellis can extract data such as account numbers, company name, utility type, meter point numbers, and usage amount in units, across various utility bill formats and terminology. This allows the company to efficiently find relevant utility data and provide energy data usage reports to their clients for strategic carbon management.

Integration:

Trellis seamlessly integrates with the company's existing systems through API endpoints. This integration enables the automatic ingestion of energy usage data from utility bills.

Data Structuring:

Trellis converts utility bill’s unstructured energy usage data into structured SQL formats.

Here are some examples of data extraction for carbon accounting use cases LLMs enable:

1. Dates: “What is the start date of the energy usage period?”

2. Utility type: “What type of utility is being used (e.g., electricity, natural gas, water)?”

3. Meter point number: “Identify the meter point number for the account.”

4. Usage amount: “How much of the utility was used at the facility?”

5. Waste type: “What type of waste was produced?”

6. Waste amount: “What is the volume or weight of the waste produced?”

Automation:

Trellis automates the carbon accounting process by extracting the requested data, conducting internal validations, and reformatting it for easy integration into existing enterprise resource planning software. This reduces the need for manual intervention and minimizes the risk of errors.

Results:

1. Scalability: By utilizing Trellis, the company can automatically process hundreds of utility bills in a day. This increase in data processing capacity empowers them to achieve an x% growth in their client base, leading to significant carbon reductions across a wider range of businesses.

2. Accuracy: Automated data extraction reduces human errors and improves energy usage accuracy by 90%, resulting in more precise carbon footprint calculations.

3. Real-Time Insights: Efficient utility data processing enables businesses to quickly develop effective carbon management strategies, helping them meet their sustainability goals and regulatory requirements.

Conclusion

Trellis helps the company provide accurate and timely carbon reports to support their clients’ sustainability goals. Through Trellis’s advanced AI capabilities, manual carbon accounting processes become streamlined by converting unstructured utility data into structured SQL format. The SQL format is easily integrated into the company's API, making energy usage data readily accessible for calculations and enabling the analysis of other relevant information such as dates, account numbers, and meter point numbers. Trellis enhances the company’s mission of advancing sustainability efforts by enabling effective carbon management.