kitewing  ·  agentic data analysis  ·  typ kw 1

AI Data Analysis. Governed. Measured. Accurate.

typ kw 1 · 4 programme · agent streaming

Kitewing is an enterprise-grade data analysis agent. It comes with centrally governed, permissioned connectors to your data sources. It features staging environments, golden prompts and unit tests for accurate answers. It deploys into our cloud or yours.

mailing list
rows analyzed today
investigations filed

reply card — one letter when we ship something worth reading. no noise.

fig. — das lochband · governed, centrally managed connections to every system you run.

Kitewing connects to Athena, BigQuery, Databricks, DynamoDB, Firestore, PostgreSQL, Redshift, Snowflake, Airtable, Google Sheets, Datadog, Looker, PostHog, App Store Connect, Apple Ads, Customer.io, Meta Business, Metricool, QuickBooks Online, RevenueCat, Salesforce, GitHub, GitLab, Notion, Slack Bot.

standard equipment · governance

Don't hope for accuracy. Guarantee it.

Kitewing's system is designed so you can measure accuracy, test it, and improve it. Deploy an answer bot to thousands of teammates with full confidence in the correctness of the system.

Golden investigations: approved question templates with quality scores tracked on every run
Audit logs: every query, source and operation on the record, filterable by workspace, member and investigation
Rubric grading: every investigation scored for data, answer and visualization quality before release
Staging: changes rehearsed against the golden investigations, scores compared live versus staging before publishing
Role-based skills: reusable analysis playbooks the agent reads on demand, assigned per group with editing scoped to admins and every edit change-tracked for review
dunkelhell

data lineage

Know where every number came from.

Every chart, graph and table in Kitewing knows the code that made it, the data sources that originated it, and the prompt that requested it. Never wonder how an agent got a number again.

Kitewing's schema usage page: a treemap of every warehouse, catalog, schema and table the agent read this month — Databricks catalogs and a Postgres database — with the top tables ranked by how many investigations touched themKitewing's data lineage modal with the Monthly active users SQL dataset selected: shape, columns, the table it reads, the question asked, and the result rowsThe Revenue by region node's detail card in Kitewing's data lineage modal, close up: the Python badge, eight rows, the question asked, and the pandas code that merged its two inputsKitewing's data lineage modal as it opens from a dashboard chart's Built from line: the Quarterly Revenue by Region chart on top, the Built from tree beneath it running down to the connectors, the output node selected
readout
nrstagetyprows
01sourcewarehouse—
02querysql6 × 2
03transformpython8
04chartdashboard—
signalweg

stage 4/4 · chart · quarterly revenue by regionEvery chart and table carries its Built from line. Press it and the trace opens beneath the chart.

more · lineage in detail →

flexible deployment

Our cloud or yours.

Run as many isolated instances in as many secure clouds as your business needs. Data in your cloud never leaves your cloud.

schematic · typ kw 1mode · saas / cloud / multiour cloudyour cloud · euyour cloud · ustyp kw 1 · kitewing environmentone environment. any cloud.
betriebsart