GreenPath SME Copilot

The ESG Assessment Flow
Evidence Library & Extraction
The Main Dashboard
The Final Sustainability Report
The AI Copilot in Action

Inspiration

The project is inspired by the unique challenges faced by Small and Medium Enterprises (SMEs) in the Vietnam/ASEAN region. These businesses often lack the specialized technical knowledge or resources required to navigate complex ESG (Environmental, Social, and Governance) reporting frameworks, such as the VSME (Voluntary SME) standard.

What it does

AI-Powered Guidance: Provides an interactive ESG Copilot that helps users understand sustainability metrics and calculation formulas (e.g., Activity Data x Emission Factor).
Structured Assessment: Guides users through an ESG assessment across Environment, Social, and Governance categories, tracking progress and status for each metric.
Automated Evidence Extraction: Uses AI to scan uploaded documents (like electricity bills or HR reports) and automatically extract relevant data, such as kWh consumption or employee headcount.
Reporting & Visualization: Generates draft sustainability reports featuring automated GHG emission trend charts and executive summaries.
Strategic Planning: Creates an actionable plan with initiatives to improve ESG performance, categorized by impact and effort.

How we built it
Frontend: Developed as a single-page application using React 19 and TypeScript for robust type safety.
Styling: Built with Tailwind CSS for a responsive, modern interface and Lucide React for iconography.
AI Integration: Integrated Google's Gemini API (specifically the gemini-3-flash-preview model) using the @google/genai SDK for natural language assistance and data extraction.
Data Visualization: Utilized Recharts to render dynamic GHG emissions trends and completion status charts.
Tooling: Managed with Vite for fast development and build cycles.

Challenges we ran into
Document Parsing: Simulating the extraction of specific data points from diverse file types (PDFs, spreadsheets, images) while maintaining a high confidence score.
Standard Alignment: Ensuring the assessment questions and reporting narrative strictly align with the VSME Basic Module requirements.
Contextual Awareness: Managing the state so the AI Copilot understands exactly which ESG topic or question the user is currently focused on to provide relevant help.

Accomplishments that we're proud of
End-to-End Workflow: Creating a seamless flow from raw document upload to a generated sustainability report.
AI Integration: Implementing a contextual AI drawer that acts as a real-time consultant for non-expert users.
Standardization: Successfully mapping complex ESG topics like Scope 1 and Scope 2 emissions into simplified, SME-friendly inputs.

What we learned
ESG Complexity: Gained a deeper understanding of the VSME standard and the specific data points required for legitimate sustainability reporting.
AI Orchestration: Learned how to ground AI responses using system instructions and real-time user context (e.g., page location and selected questions).
User Centricity: Recognized the importance of simplifying technical jargon for business owners who are not sustainability experts.

What's next for GreenPath SME Copilot
Live Document Processing: Moving beyond simulated extraction to full, real-time PDF and image processing for all evidence types.
Regional Localization: Expanding support for more localized emission factors and languages specific to various ASEAN countries.
Integration Ecosystem: Developing API connections to utility providers and HR software to automate data collection even further.
Verified Reporting: Implementing features to facilitate third-party verification of the generated reports.