Scaling Sustainability Impact: How AI Accelerates Environmental Assessments

Solving the Unstructured Data Challenge in CSRD and Scope 3 Reporting

The "dirty secret" of software is that it's just a digital form. Humans still do the heavy lifting: extracting data from messy invoices, supplier PDFs, and complex ERP reports. In this session, Josine oude Lohuis, CPO and Co-founder of Spirefly (formerly Link Nature), explains why traditional software has failed to keep up with 2026 global targets and how a new era of AI is changing the game.

"AI is changing the way we build solutions. There is an incredible gap between our global ambitions and the rate at which companies are progressing. We believe AI is the missing piece of that puzzle."
Josine oude Lohuis

Automating Environmental Assessments: From Raw Data to Impact

We’ve moved past simple chatbots. To reach your sustainability goals, you need a system that understands context and handles the math.
‍

• Phase 1: Smarter Data ExtractionLLMs now understand the meaning of a sentence. Whether it’s "200kg of recycled cotton" in an invoice or a complex supplier report, Spirefly extracts exactly what you need for an auditable assessment.
  ‍
• Phase 2: Agentic Workflows & Specialized ToolsLLMs are great at language but poor at math. We solve this by giving our AI agents specialized tools—Python scripts, Excel formulas, and regulatory frameworks—to ensure your footprints are traceable and scientifically sound.
  ‍
• Phase 3: The Human-in-the-Loop Meet our expert verification layer. We show you how AI does the "loops" of work while humans provide the final, high-value validation.
  ‍

Why Static Dashboards Are a "One-Size-Fits-No-One" Solution

Traditional dashboards don't help you make decisions. They tell you what happened, but not why it matters.
‍Watch the Spirefly Demo to see:
‍

• The Workbench: A goal-oriented space where you set objectives (e.g., "Complete CSRD by 2027") and the AI builds your plan.
  ‍
• Contextual Insights: How to teach the AI about your competitors and specific water-stress targets to get personalized advice.
  ‍
• Nature Strategy: A live look at moving from carbon data to biodiversity and land-use assessments.
  ‍

Expert Tips for Implementing Responsible AI in Sustainability Workflows

Josine shares her top "home setup" tips for sustainability leaders:
‍

• Treat AI like a new colleague: Give it context. Tell it your role, your targets, and your "why."
  ‍
• Split the responsibilities: Use AI for text-heavy tasks, but let it use specialized tools for calculations.
  ‍
• Iterate and Improve: Ask your system to check its own work. Your AI is more "agentic" than you think.

‍