About

I'm Dan Graf, CEO and co-founder of Earthchain, based in York. For more than two decades I've built software that has to be right: first the systems that authorise card payments for banks, now the systems that turn everyday business data into emissions numbers people can rely on.

Earthchain

I co-founded Earthchain in 2022. We build an AI-native sustainability data platform. Computer vision and AI workflows extract activity and consumption data from invoices, utility bills, specifications and freight records. Agents request missing evidence from the people who hold it, check what comes back and chase the gaps. Every figure stays linked to its source document, so the annual report and the operational view come from the same dataset. More on the platform.

Payments

Most of my career has been in payments technology:

• Infraxis AG, 2009–2021, Zürich and Leeds. Mission-critical software powering electronic payments for major international financial institutions, running 24/7/365. As Head of Product Development (2013–2021) I set the strategic direction for product development, planned and resourced sprints across customer projects and the product roadmap, and ran a Scrum-based process across teams in Switzerland and the UK. Before that, as QA Manager (2009–2012), I led QA and support services, consulted on payments and consumer-lending applications, and designed and built IQS, our collaborative payments testing system.
• Credit Suisse, 2007–2009, Horgen. IT project manager and solution architect: software architecture, platform engineering, and release planning and management.
• ACI Worldwide, 2002–2007, Interlaken. Technical consultant on complex card authorisation systems: requirements, engineering, certification, testing and integration.

In 2023 I gave the Day 1 keynote at Merchant Payments Ecosystem in Berlin, on sustainability and payments (watch it here).

That background shapes how I think about sustainability data. Most of it starts life as a transaction, and the disciplines of payments (reconciliation, enrichment, audit trails, never losing a record) carry straight across.

How I work

• Close to the technology. I still build things myself, from Go services to simulator plugins and boat electronics (see the work page), so my advice comes from practice.
• Product first. Technology choices follow from the customer problem and the business model, not the other way round.
• Practical AI. AI should quietly do the painstaking work inside the workflow, with declared rules, provenance and people in charge of the judgement calls.

Off the clock

I'm restoring Godiva, a 1977 Hallberg-Rassy, and sail her single-handed out of Hartlepool. I grew up skiing in the Swiss Alps, and I've built my own skis and snowboards. I studied at the University of Sussex.