Chart-Er project

Prompt-to-chart for archaeologists and researchers.

Upload CSVs, describe the chart, and get clean Vega Lite specs with sensible defaults. Built to show how AI can assist—never replace—their workflow.

View GitHub repo
Chart-Er sample chart preview
Live preview Vega Lite + Supabase + FlutterFlow

Chart-Er

Chart-Er began as a conversation with Dr. Benjamin Irvine at Bilkent University: could AI become a friendly partner for archaeologists? The result is a tool that transforms CSV uploads into charts with natural-language prompts, showing researchers how automation can support—not replace—their practice.

Under the hood, Chart-Er pairs a no-code UI with a reliable data backbone. FlutterFlow handles the experience, Supabase stores and serves data, n8n orchestrates workflows, Vega Lite generates spec sheets, and OpenAI (GPT-4.1) guides the chart logic. It’s a lightweight proof that AI can make data visualization approachable for any researcher.

  • Supabase: database + storage bucket
  • FlutterFlow: primary UI/UX
  • n8n: self-hosted workflow automation
  • Vega Lite: chart spec generation
  • OpenAI: GPT-4.1 for NL-to-chart intent
Chart-Er logo artwork

Logo concept for Chart-Er, kept beside the overview.

Sample outputs

Browse representative charts; each slide updates the explanation on the right.

Chart-Er example chart

Current slide

Prompt-driven bar chart

Quick comparison chart generated from a CSV with a short natural-language request, optimized for readability.

Slide 1 of 7