Azvai is a small circular economy practice. We use AI to do research and data work at a depth that would otherwise take a much larger team. This page explains what that means in practice for the articles on this site, and where the human work stays.
What AI does in our work
- Research. We use Anthropic’s Claude, mostly through Claude Code, to read long reports, regulations and datasets, and to pull out what is relevant. A 340-page World Bank report or a full EU regulation becomes a set of findings we can check, with page and article references.
- Data analysis. Claude writes and runs the code that pulls data straight from the source, such as the Eurostat API, the Anthropic Economic Index files or Google Search Console. The numbers in our articles come from that code, not from memory, so every figure can be traced and reproduced.
- Charts and tables. Charts are generated in Python from the same data, in a consistent house style. Where a chart carries an important number, the data table sits next to it.
- Drafting. Claude helps draft and restructure text. Borja Blanco sets the angle, rewrites and edits the result, and decides what is published.
- Publishing checks. Before anything goes live, automated checks look for placeholder text, links that do not resolve, legal references that do not point to the official source, and phrasing that breaks our style rules. Anything that fails is fixed before publishing.
What stays human
- Judgement. Which questions are worth asking, which sources to trust, and what a finding means for a business or a city come from ten years of field experience in circular economy projects.
- Verification. Numbers, laws and company claims are checked against primary sources (official statistics, EUR-Lex, BOE, company reports). If a claim cannot be sourced, it is cut.
- Responsibility. Borja Blanco reviews and edits every article and is responsible for its content. If you find an error, tell us and we will fix it.
Why we say this openly
Readers should know how a piece of content was produced. The EU AI Act sets transparency rules for AI-generated text published on matters of public interest, and we would rather be clear about our process than leave readers guessing. It is also the work we offer clients. If you want research and data workflows like these inside your own organisation, see AI for sustainability research.