Try AI data analysis on real data, no signup

The Contextflo playground is a live workspace you can question in plain English. An AI agent writes and runs the SQL against a real warehouse, using the context layer your team would build, so answers follow agreed definitions instead of guessing.

Datasets you can ask about

Rungear

A running brand: Shopify orders, ad spend, Klaviyo email and returns. New data every day.

Which marketing campaigns actually made money?
Is our return on ad spend any good?
How do online, retail and popup sales compare?
Which emails drive the most revenue?
Why are customers returning products?

Recreation.gov

Every Recreation.gov booking since FY2021, 50M rows across 11 federal agencies.

Which campgrounds made the most revenue last year?
How far ahead do people book, and does it differ by agency?
Which months are busiest for camping?
Which states have the most reservations?
How has booking revenue changed since 2021?

SF 311

Every San Francisco 311 complaint since 2008, updated daily from DataSF.

Which neighborhoods file the most 311 complaints this year?
Has graffiti gotten better or worse since 2023?
How long does the city take to close a case?
How many closed cases were actually fixed?
What are people complaining about most right now?

Questions

Do I need an account to try it?

No. Pick a dataset and ask a question in plain English. You only sign up when you want to connect your own data.

How does it know which numbers are right?

Each dataset has a context layer: table descriptions, metric definitions and saved queries that record the traps in the data, such as blended versus campaign return on ad spend. The agent reads it before writing SQL.

Can I use it on my own data?

Yes. Contextflo connects to BigQuery, Snowflake, Postgres and other warehouses, and to Google Sheets and CSV files. Your team asks questions from Claude or ChatGPT.

Hello