Pvt — From Spreadsheet to Published Data Story

Three steps to a published story

From raw data to a shareable link in under a minute.

1

Connect your data

Paste a Google Sheets URL, upload a CSV, or type data inline. Pvt keeps a live connection—your charts update when your spreadsheet does.

2

Write it, or have AI write it

The DSL is short and declarative—a few lines per chart. Write it yourself, or paste the schema into Claude, ChatGPT, or Cursor and get a working chart back on the first try.

3

Publish and share

One click to publish. Choose public, unlisted, password-protected, or domain-restricted access. Track who views your work with built-in analytics.

Why a text language beats a code notebook

Pvt uses a concise markup language instead of Python or JavaScript. Here's why that matters.

Readable

You can read the chart without running it

Five lines tell you exactly what a chart does: what data, what axes, how it's grouped, how it's styled. No scrolling through 40 lines of matplotlib boilerplate.

Pvt

series: column(stack)
x: Quarter
y: Revenue
groupby: Region
format: legend-right

5 lines

Python

import pandas as pd
import matplotlib.pyplot as plt

df = pd.read_csv('revenue.csv')
pivot = df.pivot_table(
  values='Revenue',
  index='Quarter',
  columns='Region',
  aggfunc='sum'
)
ax = pivot.plot(
  kind='bar', stacked=True,
  figsize=(10, 6)
)
ax.legend(loc='center right')
plt.title('Revenue by Region')
plt.tight_layout()
plt.show()

18 lines

LLM artifact

One chart, one file
No URL to share
No version history
No viewer analytics

Pvt document

Charts, tables, slides in one doc
Published URL: pvt.app/@you/story
Version history + rollback
Viewer analytics + tracked sharing

Beyond artifacts

An artifact is a chart. Pvt is a publishable story.

Ask an LLM for a chart and you get a single-file artifact—great for a one-off question, useless for an exec review. Pvt's DSL composes charts, dashboards, and slides into one document that publishes to its own URL—with version history and viewer analytics built in. Same AI. Finished product.

Zero environment

No install. No runtime. No errors.

No pip install. No Jupyter. No version conflicts. No “works on my machine.” Open your browser, connect your data, get a chart. It runs everywhere because there's nothing to run.

$ pip install pandas matplotlib seaborn
$ jupyter notebook
$ # fix numpy version conflict...
$ # google "matplotlib tight_layout warning"
$ # export to PNG, email it, lose interactivity

None of this.

Four ways to share one document

Write it once. Present it however your audience needs it—no copy-paste, no re-export.

Interactive Charts

40+ chart types powered by Highcharts. Tooltips, zoom, drill-down, and responsive design by default. Line, bar, scatter, heatmap, treemap, sankey, gauge, and more.

Browse chart types →

$2.4M

Revenue

+18%

Growth

1,205

Customers

Live Dashboards

KPI modules, sparklines, data tables with conditional formatting, and drill-down hierarchies. Connected to Google Sheets for real-time updates.

See dashboard examples →

Slide Presentations

Full-screen presentations with speaker notes, transitions, multi-column layouts, and embedded live charts. Present from your browser—no PowerPoint needed.

See presentation examples →

pvt.app/s/your-story

Published Pages

Real web pages with their own URLs. Public, unlisted, password-protected, or domain-restricted. Version history and viewer analytics built in.

Learn about publishing →

Built for the space between spreadsheets and code

If you've tried these tools, you know the gap Pvt fills.

	Pvt	Google Charts	Python / Jupyter	Hex / Observable
Learning curve	DSL in minutes, or let AI write it	Point and click	Python required	SQL + Python
Chart types	40+ interactive	~15 basic	Unlimited (manual)	Good selection
Live data connection	✓	✓	✗	✓
Dashboards + KPIs	✓	✗	With frameworks	✓
Slide presentations	✓	✗	✗	✗
Publishing + access control	✓	Embed only	Manual hosting	✓
No install required	✓	✓	✗	✓