Earth5R’s Agentic Approach to Data-Driven Climate Action

A 50,000-kilometre cycling journey across Asia, Africa, and Europe exposed Saurabh Gupta to the scale of environmental destruction and the widespread indifference surrounding it. On returning home, he founded Earth5R to enable climate action through data-driven sustainability programs, connecting governments, industries, researchers, and communities into one integrated impact network.

Over 11+ years of implementing sustainability programs at scale, Earth5R identified a critical gap: sustainability decisions are often driven by abstract, internet-derived data that fails to reflect on-ground realities. Governments and organizations need intelligence that is implementation-aware, data-backed, and context-sensitive. Earth5R addresses this gap by transforming real-world environmental action and community-generated data into structured, AI-driven sustainability intelligence.

Earth5R had accumulated a vast volume of high-quality field data ranging from reports, surveys, images, and videos to case studies across geographies and themes. However, turning this rich, unstructured data into actionable insights for internal teams and external stakeholders was time-intensive and difficult to scale.

Through the AI Agents Masterclass in association with Google for Startups, Earth5R gained knowledge and insights to develop an Agent which operates through a structured, multi-stage workflow:

1. Identifies the sustainability theme or question

2. Retrieves relevant internal Earth5R datasets

3. Maps this internal intelligence against external policy documents and scientific literature

4. Synthesizes outputs as articles or explainers based on tested, on-ground practices

The system is designed as a multi-step, agentic workflow where each component has a clearly defined role ensuring stability, transparency and control. Built using n8n, Google Sheets, AI nodes, and WordPress integrations, the Agent functions as a connected system rather than a standalone model.

The Agent is currently supporting Earth5R’s internal teams with research and content creation. Early feedback highlights improved depth, contextual accuracy and credibility. The next phase involves controlled external pilots, with early discussions underway with organizations including ICICI Bank, Swiss Air, IITs, and NYU.