Introduction to Mapping Agent

What is Mapping Agent

Calculating emissions requires matching each business activity, such as a purchase, a shipment, an energy bill to the correct emission factor. When you are working with large volumes of data, doing this manually is slow and inconsistent.

Climatiq’s AI-powered Mapping Agent (formerly known as Autopilot) automatically maps your business activity data with relevant emission factors. It automates this step by taking a plain-language description of your business activity, for example, “grid electricity”, “business class flight”, or “polypropylene packaging film” and searching Climatiq’s database to find the best-matching emission factors. Each result comes with a quality label that tells you how confident the model is in the match, so you know which results to use directly and which ones to check first.

How it works

Mapping Agent matches your business activity descriptions against Climatiq’s emission factor database and evaluates each candidate emission factor for relevance. The degree of relevance is determined internally by the AI model and may shift as the model improves over time. Each result is assigned a quality label (accept or review) based on how strong the match is. The label gives you a stable, actionable signal regardless of how the underlying model evolves.

Mapping Agent also reasons about each match: rather than purely scoring descriptions against factors, it evaluates whether an emission factor genuinely represents your activity. If the top results are weak or ambiguous, Mapping Agent rephrases your input and searches again, repeating until it finds strong matches or reaches the allowed number of attempts.

By default, Mapping Agent prioritizes accuracy over speed. If speed matters more than getting the highest possible match quality, it can be configured to run faster with some trade-off in accuracy. See the Mapping Agent API reference for configuration details.

Example: For the input “oat milk latte” Mapping Agent returns a ranked list of matched emission factors. The top result is a plant-based dairy alternative or food and beverage production factor, labeled accept, meaning you could use it in your calculation. Lower-ranked alternatives are labeled review, flagging them for a manual check before use. Depending on your mode (see below), you might be able to accept the top result automatically or inspect the list before selecting.

When to use it

Climatiq has dedicated calculation features for specific activity types, including Travel, Energy, Freight, and Procurement, that handle structured inputs and return CO2e directly. Use those when your data fits a supported activity type.

Mapping Agent is the right choice when:

• Your data is unstructured or inconsistently labeled: invoice descriptions, supplier names, spend categories, or any free-text inputs that do not map directly to a known activity ID
• Your activity type does not have a dedicated calculation feature: waste, leased assets, downstream product activities, franchises, investments, and other categories not covered by a dedicated Climatiq tool
• You are processing high volumes of varied descriptions: Mapping Agent handles diverse inputs at scale without requiring manual curation of emission factors for each line
• You want to automate factor selection: rather than searching and selecting emission factors manually for each activity

If you are unsure which emission factor applies to an activity, Mapping Agent is a reliable starting point. Submit the description and use the quality labels to assess how much confidence to place in the result.

Workflows of the Mapping Agent

Mapping Agent offers three workflows depending on how much control you need over the matching process:

Suggest: Mapping Agent returns a ranked list of the best-matching emission factors for your description. You review the options and choose which factor to use. This is the right workflow when you want a human or your application logic to make the final factor selection.

Suggest + Estimate: A two-step flow where you first get a ranked list of matched factors from Suggest (each result includes an activity_id), then use that activity_id with Basic Estimate to receive a CO2e calculation. This gives you control over factor selection while still automating the final calculation.

One-shot Estimate: Mapping Agent automatically selects the number 1 top-ranked match from Suggest, then calls Estimate using that suggestion, and returns a CO2e calculation in a single step. No selection required. This is the right workflow when you want fully automated processing at scale and are comfortable letting Mapping Agent make the final call.

Where Mapping Agent is available

Mapping Agent powers emission factor matching across Climatiq’s product suite:

• Climatiq API: Access Mapping Agent directly via the /mapping-agent/v1 endpoint
• Data Studio : Mapping Agent maps your activity data to emission factors automatically when using the AI Matching bulk calculator
• PCF Studio : Mapping Agent matches component and material descriptions to emission factors when calculating product carbon footprints
• Excel Add-in: Mapping Agent powers factor matching via custom Excel formulas, directly within your Excel environment
• Google Sheets Extension: Mapping Agent is available for factor matching directly within your spreadsheet

The specific version of Mapping Agent available in each product may vary. Check the relevant product documentation for the latest details.

Supporting resources