When someone asks an AI what's on in Athens tonight, why isn't there a source built to answer? I built one. Agent Athens — an AI-native events platform for Athens, built end to end.
Agent Athens is over 3,800 static pages, generated daily and unattended: 10 scrapers feed a SQLite store, an AI enrichment layer rewrites raw data into editor-quality copy, and the whole thing redeploys on a schedule with zero human touch — and zero external API cost, because the enrichment runs inside Claude rather than calling out to one.
The architecture is the actual bet. Instead of chasing traditional SEO, the system is engineered to be cited by answer engines — ChatGPT, Perplexity, Claude — as the authoritative source on Athens culture: structured data, entity linking, enrichment that reads like a person wrote it. It's a working argument about where discovery is going, not just an events list.
Open to AI roles — automation, or putting LLMs to work on real problems. I prefer elegant solutions and ship scrappy ones when it counts.
Download CV