Making sustainability easier to prove starts with asking what we need to measure. Carbon matters, but so do water use, land use, and the wider effects of the things we build. That was on my mind during CarbonGraph’s recent team gathering in Vancouver, and it has stayed with me as I think about where we take the product next.

What can trees teach us about infrastructure?

Our visit to Capilano, alongside Alex’s “Trees as Infrastructure” theme, gave me a useful way to step back from the day-to-day work of building a product.

Infrastructure shapes how we live and work. It also sits within natural systems that we depend on. Thinking about trees as infrastructure brought those connections into focus for me: how we use land, how we use water, and how we design for the long term.

It was a reminder to keep those questions in view when we talk about sustainable design. A carbon footprint answers an important question. Understanding environmental impact asks us to look more broadly, and to consider the trade-offs behind a decision.

As head of product, I came back to a practical question: how do we help more people provide the evidence behind their sustainability decisions, sooner?

For CarbonGraph, I see two connected opportunities.

How do we make our calculation engine easier to access?

Over the last five years, we have built the calculation engine at the heart of CarbonGraph. We want more people to be able to put that power to work in the ways that fit their organizations.

For a partner serving many manufacturers, that could mean connecting environmental calculations with the systems and workflows they already use. For a smaller company facing its first request for an Environmental Product Declaration (EPD), it could mean a more approachable way to get started with the data and analysis it needs.

We are seriously considering building a Model Context Protocol (MCP) server. MCP is an open standard for connecting AI applications to external tools and data. A CarbonGraph MCP server could let companies using their own AI agents work with our calculation capabilities through those agents. Learn more about MCP.

That is a direction we are exploring. The question guiding it is straightforward: how can we make credible environmental calculations easier to access, so partners and manufacturers can spend less effort connecting the pieces and more time using the results?

How do we bring AI into the work of modelers, analysts, and verifiers?

Over the past six months, we have been working to make CarbonGraph an AI-native company in how we develop software. Our Vancouver discussions gave us time to consider what that experience could mean for the people using CarbonGraph.

Where could AI help a modeler move through repetitive work? What would make an analyst’s next step clearer? How could it help a verifier find and understand the evidence behind a result?

Those are product questions, and they are also questions about trust. We need to think carefully about access to sensitive information, the actions an AI system is allowed to take, and where a person needs to review its work. A faster workflow still needs assumptions and calculations that people can examine.

For a small team without dedicated life cycle assessment (LCA) expertise, a more approachable experience could make the first steps less daunting. For experienced practitioners, thoughtful AI support could free more attention for the judgments that need their expertise.

What I am taking forward from Vancouver

Capilano reminded me why the broader picture matters. The products and infrastructure we build have effects beyond a single carbon number. People need ways to understand those effects and substantiate the choices they make.

My focus is on accelerating that proof: opening up access to the calculation power we have built, and exploring AI support that helps people use it thoughtfully and securely.

If you are a partner working with AI agents, or a manufacturer trying to make sense of your next EPD request, I would like to hear where you get stuck. Those are the problems we want to build around.