Climative builds low carbon plans
Climative's responsible AI framework

Responsible AI Framework

For Virtual Home Energy Assessments

A home is more than just a financial asset. It’s where we rest, play, work, and raise our families. As stewards of home energy data, the Climative team recognizes the tremendous responsibility of ensuring this data is used responsibly to generate insights that move us toward a net-zero future.

Responsible AI Framework for Virtual Home Energy Assessments

Climative’s Responsible AI Framework for Virtual Home Energy Assessments ensures that our team is deploying next-generation technologies with ethics, equity, and transparency in mind. The five principles of this Framework support our mission to re-invent how organizations collaborate and empower all building owners to create great spaces to live and work in the planet’s best interest.

Click to Expand

Dig Into the 5 Principles of Climative's Responsible AI Framework

Click to expand each section that provides more detail on the five principles of Climative’s Responsible AI Framework.

As AI is increasingly adopted for decision-making tasks across industries and within the public sector, it’s more important than ever that these technologies are developed with robust risk management strategies and ethical standards.  

Climative’s AI process is a comprehensive approach designed to solve business problems using artificial intelligence and machine learning and to manage the risks associated with emerging technology. The steps involved in developing, testing, deploying, and monitoring our models are as follows:  

  • Business Understanding (Understanding the problem to be solved) 
  • Data Intake from Trusted Sources 
  • Data Cleaning and Augmentation 
  • Model Design
  • Model Tuning and Building 
  • Model Evaluation Using Internal Standards 
  • Industry Standards Compliance 
  • Explainable AI 
  • Model Deployment and Integration 
  • Continuous Testing and Monitoring


Climative Insights employs an AI-assisted approach to assess the housing stock’s energy efficiency and carbon emissions at scale. We prioritize accuracy, consistency, fairness, and transparency by integrating an AI process into our modeling procedure. This approach empowers homeowners, energy efficiency organizations, utilities, governments, real estate organizations, and financial institutions to make informed decisions to reduce carbon emissions from residential buildings.

Click here to visualize the AI process.

Climative’s AI-assisted energy insights adhere to industry standards established by trusted institutions. This provides several benefits for Climative, its partners, and its users:

  • Interoperability: enables seamless collaboration with services from other vendors like governments, financial institutions, insurance providers, and real estate professionals.
  • Quality assurance: products and services meet internal and industry standards, providing a benchmark for reliability and performance.
  • Compatibility: field-level compatibility with leading energy assessment software in North America, simplifying integration and fostering a more efficient product and service ecosystem.
  • Regulatory Compliance: helps customers and partners meet regulatory requirements, ensuring safety, security, and ethical practices.
  • Customer Trust: following established standards enhances the credibility of Climative, its customers, and partners.
  • Risk Management: Standards offer guidelines and best practices, helping Climative manage risks by reducing errors, security vulnerabilities, and other potential issues.


We use field-level compatibility with current home energy modeling standards to ensure maximum integration with programs like EnerGuide and EnergyPlus. Additionally, we have adopted the Framework outlined in ASHRAE Guideline 14 to establish a methodology for assessing the accuracy of our predictions relative to onsite audit results. Climative is actively collaborating with government agencies, standards organizations, and industry working groups to promote standardization in digital home energy assessments. The Climative data platform continually improves to accommodate new data sets and standards.

Various third parties have audited our virtual home energy assessment methodology and results, and it has proved to have high central tendency and low dispersion versus onsite audit results, as well as high consistency across different regions and building archetypes. We invite you to book a call with a team member to discuss the results of these third-party audits.

Climative’s goal is to empower homeowners to take impactful retrofit actions for a home that is comfortable, affordable, and planet-friendly. Just as a home is an ever-changing space, our energy insights are a living record. Climative’s virtual home energy assessment gives homeowners control over their insights in several ways throughout their retrofit journey: 

  • Homeowners can fill out questionnaires about their homes to improve the fidelity of their insights and recommendations. 
  • Homeowners can grant access to incorporate other trusted data sets, such as energy use data and results from on-site or remote energy assessments. 
  • Climative low-carbon plans offer several retrofit recommendations, allowing homeowners to make retrofits that fit their lifestyle and budgets. 
  • Home energy retrofit recommendations include very detailed breakdowns of energy use, emissions impact, and cost, helping homeowners make informed decisions. 

Climative’s energy assessments continuously improve with the availability of more data, ensuring high-quality and accurate results. The confidence score associated with each energy assessment report is transparent to both homeowners and our customers. Actionable and responsible recommendations are made so that different levels of the energy report can be leveraged for various use cases. 

Climative offers three levels of virtual home energy assessments: 

  1. No-touch mass market home energy assessment (level 1): This leverages public data and serves as a starting point for homeowners to understand and plan impactful retrofits. Progressive governments and energy efficiency organizations have used level 1 assessments to aggregate data for informed decisions on policies and programs that increase participation and drive high-impact home retrofits.
  2. Survey-based assessment (level 2): This assessment gathers data from homeowners or real estate agents through online questionnaires, providing a more comprehensive understanding of the building when combined with data from a level 1 assessment. It enhances a region’s overall database for tracking retrofit and carbon reduction.
  3. Remote home energy assessment (level 3): This comprehensive evaluation is conducted via phone or web conference with a professional energy advisor. It provides homeowners with a detailed report outlining energy-saving and carbon-reducing recommendations, estimated cost-savings, and coaching on potential retrofit options. 

Learn more about the three levels of virtual home energy assessments in our whitepaper “Scaling Net-Zero Retrofits.” 

We believe technology should be used to create prosperity and opportunity for all, including those groups who have been historically underserved. That’s why we deploy our AI-assisted assessments to be accessible, with 100% coverage of a region’s residential building stock, regardless of location, income, or demographics.  

On-site energy assessments, a popular tool for evaluating homes and unlocking financial support for homeowners, are an educational and engaging activity that comes with some challenges. These home visits can be time-consuming and costly for the evaluator, the homeowner, and taxpayers. Furthermore, folks in rural areas can’t always access this service. AI-assisted home energy assessments are convenient and affordable for the homeowner and available to anyone in the region with an Internet connection and home address. We collaborate with municipalities, utilities, and energy efficiency organizations to provide energy reports to homeowners, facilitating meaningful retrofits. These virtual energy reports are a starting point in the retrofit journey, educating homeowners before they engage with the experts and institutions. 

By design, machine learning models allow for the rapid and cost-effective collection and analysis of high-quality data. Climative’s AI models assist the creation of equitable policies and programs by reducing barriers to information and incentives. This digital and regional approach successfully facilitates building retrofits for communities with the highest barriers to participation, e.g., low-income families and renters. With building-level home energy data and retrofit recommendations, municipalities and utilities can offer tailored financial incentives and resources to the homes and neighborhoods that need them most.  

When users interact with Climative’s data platform, we believe it’s their right to know the following: 

  • How machine learning was leveraged to determine energy insights 
  • Which data sets were used to calculate a home’s low-carbon plan 
  • Energy insights that are easy to understand and actionable 
  • Climative’s analytical methods are open to inspection 

Climative leverages explainable machine learning to help users comprehend the models and trust the predictions. One aspect of our model is sensitivity analysis: a visualization of how inputs affect energy consumption, GHG emissions, and energy cost. This is used to run “what if” scenarios for governments and energy efficiency organizations to design highly successful programs. 

Although our models form Climative’s unique intellectual property, they are not positioned as a “black box.” Our partnership proposals and agreements very clearly explain the relationship between inputs and outputs, so customers have confidence in our approach. Homeowners can feel confident that their home’s energy insights are detailed and reliable and offer significant savings on energy and operating costs. 

By following the five principles outlined in our Responsible AI Framework, Climative ensures that our AI-assisted home energy assessments are not only accurate and transparent but also provide meaningful energy insights and responsible, actionable recommendations to various stakeholders, contributing to a low-carbon future. 

Tianshu Huang

Meet Our Data Scientist

Tian is a dedicated and creative statistical modeler who leads the development of Climative’s machine learning algorithms. She believes that ethical and thoughtful use of AI will help us achieve climate goals while creating equal opportunity for all.

Tianshu Huang Climative Data Scientist

Interested in Changing the trajectory of climate change?

Climative is building a climate-positive world through innovative, collaborative, and data-driven solutions.

This is your invitation to be part of the journey

Low Carbon Dashboard