Quail was an AI-native BI tool that let anyone query data in plain English.
Built by two founders. We closed it after not finding product-market fit. Here's what we built and learned.
What We Built
AI-powered SQL editor (plain English → explainable SQL), agentic hydration for fast visualizations, dynamic chart rendering with live JSX compilation, dashboards, and rich-text reporting.
How It Worked
Two-step agentic hydration: the model received a tiny sample to propose a visualization scaffold, then the client hydrated it with full data locally. Fast, cheap, and interactive.
Tech Stack
React, TypeScript, Tailwind, Radix UI, Framer Motion, Monaco Editor. Vercel AI SDK with Azure OpenAI. Chart.js, Recharts, D3. MongoDB + Supabase.
Lessons
Interest in natural-language SQL was high, but adoption required workflows around the model: governance, review, ownership, scheduling. PMF needed narrower, high-value use cases with clear ROI.