Hi Product Hunt! 👋 I'm Enrique — a computer-science professor, and the solo founder of Tagaroo. Tagaroo exists because annotation is where research quality is actually decided, and it's still mostly done with highlighters, spreadsheets, and heroic patience. I built it on a simple bet: an AI agent can do the exhausting first pass over transcripts, text, and images, while the researcher keeps every judgment call. What it does: 🏷️ The agent takes the first pass. Joey (the agent) applies your rating scale or codebook and hands back labeled spans — each with the evidence, a severity, and a written rationale. You accept, edit, or reject. 📐 Any ontology. Upload a PDF of your codebook or scale and Tagaroo turns it into a live annotation workflow. There's also a library of 25 validated clinical and behavioral scales (PHQ-9, HAM-D, MITI…) ready to apply. 👥 Reliability built in. Multi-annotator studies with inter-rater agreement computed as you work, plus a compare-and-consolidate view for disagreements. 📊 Measured, not asserted. We publish benchmarks of the agent against human raters — you can score it against your own annotators too. It's browser-based and self-serve: the free plan needs no card, and trial data never leaves your browser. Everything is EU-hosted and GDPR-first. For PH: 30% off Pro for your first 3 months with code PHUNT30 (until Sep 22). I'd genuinely love your feedback — especially: what data are you annotating today, and what would the agent have to get right before you'd trust it with a first pass? I'll be here all day answering everything.

Tagaroo

The AI agent takes the first annotation pass. You review.

The AI agent takes the first annotation pass. You review.