Skip to content
Carbon GPT Carbon GPT Carbon GPT

Using the AI Assistant

This tutorial guides you through the process of leveraging Carbon GPT's AI Assistant to gain insights, receive recommendations, and automate analysis of your emissions data.

Prerequisites

  • Completed Analyzing Emissions Data tutorial
  • Emissions data entered and calculated
  • Basic understanding of carbon accounting concepts

Tutorial Overview

In this tutorial, you will learn how to:

  1. Access and interact with the AI Assistant
  2. Ask effective questions using natural language
  3. Interpret AI-generated insights
  4. Use AI for automated data analysis
  5. Implement AI recommendations
  6. Customize the AI experience
  7. Integrate AI insights into workflows
  8. Leverage AI for reporting and compliance

Accessing the AI Assistant

Finding the AI Assistant Interface

  1. Log in to your Carbon GPT account
  2. Access the AI Assistant through:
    • The persistent chat icon in the bottom right corner
    • The dedicated AI Assistant tab in the main navigation
    • The contextual AI button within specific features

Understanding Capabilities and Limitations

CapabilitiesLimitations
Data analysis and interpretationCannot modify your data without permission
Trend identificationRequires sufficient historical data
Recommendation generationRecommendations need human review
Regulatory guidanceNot a substitute for legal advice
Report draftingFinal reports require human verification
Natural language understandingMay need clarification for complex queries

Setting Up Preferences

  1. Navigate to Settings > AI Assistant
  2. Configure your preferences:
    • Default analysis scope
    • Preferred visualization styles
    • Technical language level
    • Response detail level
    • Data privacy settings

Accessing Help and Examples

  1. Type "help" or click the Help button
  2. Browse suggested prompts by category:
    • Data Analysis
    • Reporting
    • Compliance
    • Reduction Strategies
    • Best Practices
  3. View tutorial videos and documentation

Natural Language Interactions

Formulating Effective Queries

  1. Be specific about what you're looking for:

    • Vague: "Show me my emissions"
    • Specific: "Show me Scope 1 emissions for Q2 2024 by facility"

  2. Include relevant parameters:

    • Time period
    • Emission scope
    • Organizational units
    • Comparison requests

  3. Specify output format when needed:

    • "Generate a chart showing..."
    • "Create a table comparing..."
    • "Summarize in bullet points..."

Types of Questions the AI Can Answer

Descriptive Questions

  • "What were my total emissions last quarter?"
  • "Which facility has the highest Scope 2 emissions?"
  • "How has my carbon intensity changed over the past year?"

Analytical Questions

  • "Why did my emissions increase in March?"
  • "What factors are driving my Scope 3 emissions?"
  • "How do weather patterns correlate with my energy usage?"

Predictive Questions

  • "What will my emissions be next quarter if current trends continue?"
  • "When will I reach my reduction target at the current rate?"
  • "How would implementing solar panels affect my emissions profile?"

Prescriptive Questions

  • "What actions should I prioritize to reduce emissions?"
  • "How can I improve my CDP score?"
  • "What emission factors should I use for my fleet vehicles?"

Using Natural Language Effectively

  1. Start with conversational phrases:

    • "Can you tell me..."
    • "I'd like to understand..."
    • "Help me analyze..."

  2. Use follow-up questions to refine results:

    • "Can you break that down by department?"
    • "Show me that as a percentage instead"
    • "Compare that to last year's data"

  3. Provide context when needed:

    • "Looking at our recent acquisition of Facility X..."
    • "Considering our goal to reduce emissions by 30%..."
    • "In preparation for our CDP submission..."

Sample Conversation Flow

User: What were my total emissions last quarter?
AI: Your total emissions for Q2 2024 were 1,250 tCO₂e, a 5% decrease from Q1 2024.


User: Can you break that down by scope?
AI: Here's your Q2 2024 emissions by scope:
- Scope 1: 450 tCO₂e (36%)
- Scope 2: 600 tCO₂e (48%)
- Scope 3: 200 tCO₂e (16%)


User: Why did Scope 2 decrease compared to Q1?
AI: Scope 2 emissions decreased by 15% compared to Q1 due to:
1. Increased renewable energy procurement (10% impact)
2. Seasonal reduction in cooling needs (3% impact)
3. Implementation of energy efficiency measures (2% impact)

Interpreting AI-Generated Insights

Understanding AI Responses

  1. Review the complete response:

    • Summary statements
    • Supporting data
    • Visualizations
    • Recommendations
    • Confidence indicators

  2. Identify key components:

    • Facts vs. interpretations
    • Historical data vs. projections
    • Direct answers vs. suggestions

Evaluating Confidence Levels

  1. Look for confidence indicators:

    • Explicit confidence statements
    • Qualifying language ("likely," "possibly")
    • Data quality notes
    • Alternative interpretations offered

  2. Consider data limitations:

    • Sample size
    • Data completeness
    • Historical context
    • Unusual circumstances

Verifying AI Suggestions

  1. Cross-reference with other data sources
  2. Consult domain experts when appropriate
  3. Review underlying calculations
  4. Consider business context not available to the AI

Handling Ambiguous Responses

  1. Ask clarifying questions:

    • "Can you explain what you mean by...?"
    • "What data are you basing this on?"
    • "How confident are you in this analysis?"

  2. Request alternative approaches:

    • "Is there another way to look at this?"
    • "What other factors might explain this?"
    • "Can you analyze this using a different method?"

Automated Data Analysis

Automated Trend Identification

  1. Navigate to AI Assistant > Automated Insights
  2. Select Trend Analysis
  3. Configure analysis parameters:
    • Data series to analyze
    • Time period
    • Granularity (daily, weekly, monthly)
    • Significance thresholds
  4. Review identified trends:
    • Long-term trends
    • Seasonal patterns
    • Step changes
    • Acceleration/deceleration points

Anomaly Detection

  1. Navigate to AI Assistant > Automated Insights
  2. Select Anomaly Detection
  3. Configure detection parameters:
    • Sensitivity level
    • Baseline period
    • Alert thresholds
  4. Review detected anomalies:
    • Data points outside expected ranges
    • Pattern disruptions
    • Sudden changes
    • Missing or inconsistent data

Comparative Analysis

  1. Ask the AI to compare:
    • Different time periods
    • Facilities or business units
    • Your performance vs. benchmarks
    • Actual vs. targeted performance
  2. Specify comparison metrics:
    • Absolute values
    • Percentage changes
    • Normalized metrics
    • Composite scores

Predictive Insights

  1. Request forecasts and projections:
    • "Forecast my emissions for the next 4 quarters"
    • "When will I reach my reduction target?"
    • "Project the impact of switching to renewable energy"
  2. Review prediction components:
    • Baseline projection
    • Confidence intervals
    • Key assumptions
    • Sensitivity analysis

Implementing AI Recommendations

Evaluating Recommendation Feasibility

  1. Review recommendation details:

    • Implementation requirements
    • Expected benefits
    • Potential challenges
    • Timeline considerations

  2. Consider organizational factors:

    • Budget constraints
    • Resource availability
    • Strategic alignment
    • Regulatory requirements

Prioritizing Suggested Actions

  1. Navigate to AI Assistant > Recommendations
  2. View recommendations sorted by:
    • Impact potential
    • Implementation difficulty
    • Cost-effectiveness
    • Time sensitivity
  3. Create a prioritized action list

Creating Implementation Plans

  1. For selected recommendations, click Create Plan
  2. Define implementation steps:
    • Action items
    • Responsible parties
    • Timelines
    • Resource requirements
    • Success metrics
  3. Set up progress tracking

Tracking Recommendation Outcomes

  1. Navigate to AI Assistant > Recommendation Tracking
  2. Monitor implementation status
  3. Record actual outcomes
  4. Compare results to AI predictions
  5. Provide feedback to improve future recommendations

Customizing the AI Experience

Training on Your Terminology

  1. Navigate to Settings > AI Assistant > Custom Terminology
  2. Add organization-specific terms and definitions
  3. Upload glossaries or reference documents
  4. Review and approve AI-suggested terminology

Creating Custom AI Workflows

  1. Navigate to Settings > AI Assistant > Workflows
  2. Create a new workflow:
    • Define trigger conditions
    • Configure analysis steps
    • Set output formats
    • Specify delivery methods
  3. Test and refine the workflow
  4. Schedule or trigger manually

Setting Up Regular AI Check-ins

  1. Navigate to Settings > AI Assistant > Scheduled Insights
  2. Configure regular analysis:
    • Weekly performance summaries
    • Monthly trend analysis
    • Quarterly deep dives
    • Annual review preparation
  3. Specify delivery preferences:
    • In-app notifications
    • Email reports
    • Dashboard widgets
    • Calendar invites with findings

Integrating AI Insights into Workflows

Incorporating AI in Decision-Making

  1. Request decision support:
    • "Help me evaluate these reduction options"
    • "Analyze the cost-benefit of these projects"
    • "Compare these emission factor approaches"
  2. Use AI insights as input for:
    • Investment decisions
    • Strategy development
    • Performance reviews
    • Goal setting

Sharing AI Insights with Stakeholders

  1. Generate shareable content:
    • "Create a summary of this analysis for my team"
    • "Prepare an executive briefing on these findings"
    • "Generate slides explaining this trend"
  2. Configure sharing options:
    • Export format
    • Detail level
    • Visual elements
    • Supporting data

Automating Routine Tasks

  1. Identify repetitive analysis tasks
  2. Create AI automation:
    • Data quality checks
    • Regular report generation
    • Performance monitoring
    • Anomaly alerts
  3. Configure triggers and delivery

Leveraging AI for Reporting and Compliance

Report Drafting Assistance

  1. Request reporting help:
    • "Help me prepare my CDP report"
    • "Draft a sustainability report section on emissions"
    • "Create an executive summary of our carbon performance"
  2. Review and refine AI-generated content:
    • Verify data accuracy
    • Enhance narrative elements
    • Add organization-specific context
    • Ensure compliance with requirements

Regulatory Compliance Guidance

  1. Navigate to AI Assistant > Compliance
  2. Select relevant frameworks:
    • GHG Protocol
    • CDP
    • TCFD
    • EU CSRD
    • SEC Climate Disclosure
  3. Request compliance assessment
  4. Review guidance and recommendations

Dashboard Integration

  1. Navigate to Dashboard > Settings > AI Integration
  2. Configure AI-powered dashboard widgets:
    • Insight summaries
    • Anomaly detection alerts
    • Trend spotlights
    • Recommendation highlights
    • Compliance status updates
  3. Set refresh frequency and display preferences
  4. Position widgets on your dashboard layout

Using AI Assistant with Dashboard

  1. Click the AI icon on any dashboard widget to:
    • Ask questions about the displayed data
    • Request deeper analysis of trends shown
    • Generate alternative visualizations
    • Save insights to dashboard notes
  2. Use the "Ask about my dashboard" feature to:
    • Get a summary of all dashboard metrics
    • Identify correlations between different widgets
    • Receive suggestions for dashboard improvements
    • Generate executive summaries of dashboard data
  3. Set up AI-powered dashboard alerts:
    • Anomaly detection thresholds
    • Progress against targets
    • Data quality issues
    • Compliance deadline reminders

Customizing AI Dashboard Experience

  1. Navigate to Settings > Dashboard > AI Preferences
  2. Configure your AI dashboard experience:
    • Default insight types to display
    • Alert sensitivity and frequency
    • Technical language level
    • Visual style preferences
  3. Create role-based dashboard views with appropriate AI insights

Reduction Strategy Development

  1. Navigate to AI Assistant > Strategy
  2. Input parameters:
    • Reduction targets
    • Timeline
    • Budget constraints
    • Industry context
  3. Review strategy recommendations:
    • Short-term actions
    • Medium-term initiatives
    • Long-term transformations
    • Implementation roadmap

Example Use Cases

Emissions Analysis and Insights

  • Query: "What are my top emission sources and how have they changed over the past year?"
  • AI Response: Provides a ranked list of emission sources, percentage contribution to total, year-over-year changes, and contributing factors to significant changes.

Performance Benchmarking

  • Query: "How do my emissions compare to industry benchmarks for a retail company of our size?"
  • AI Response: Compares your emissions to industry averages, identifies areas where you outperform or underperform, and suggests focus areas for improvement.

Reduction Opportunity Identification

  • Query: "What reduction opportunities should I prioritize based on our current emissions profile?"
  • AI Response: Analyzes your emission sources, identifies high-impact reduction opportunities, estimates potential savings, and suggests implementation approaches.

Report Generation

  • Query: "Generate a summary of my Q1 performance for the executive team"
  • AI Response: Creates an executive summary with key metrics, significant changes, progress against targets, and strategic recommendations.

Compliance Assistance

  • Query: "Help me prepare for CDP reporting by identifying gaps in my current data"
  • AI Response: Reviews your data against CDP requirements, identifies missing elements, suggests data collection approaches, and provides reporting tips.

Advanced AI Features

Custom Analysis Requests

  1. Navigate to AI Assistant > Custom Analysis
  2. Define analysis parameters:
    • Data sources
    • Analysis objectives
    • Preferred methodologies
    • Output format
  3. Submit and review results

Report Generation Assistance

  1. Navigate to AI Assistant > Report Builder
  2. Select report type
  3. Configure content parameters:
    • Sections to include
    • Data range
    • Detail level
    • Target audience
  4. Generate draft and refine

Regulatory Compliance Guidance

  1. Navigate to AI Assistant > Compliance
  2. Select relevant frameworks:
    • GHG Protocol
    • CDP
    • TCFD
    • EU CSRD
    • SEC Climate Disclosure
  3. Request compliance assessment
  4. Review guidance and recommendations

Reduction Strategy Development

  1. Navigate to AI Assistant > Strategy
  2. Input parameters:
    • Reduction targets
    • Timeline
    • Budget constraints
    • Industry context
  3. Review strategy recommendations:
    • Short-term actions
    • Medium-term initiatives
    • Long-term transformations
    • Implementation roadmap

Next Steps