Welcome to Kelvin!

Kelvin is a datification engine - a search engine that continuously refines billions of unstructured documents into accessible data sets. Kelvin is designed for economists, investment professionals, and machine learning processes in need of unique financial time series derived from big data. Kelvin's data can be downloaded in bulk, streamed in real time directly to your application, or analyzed in the cloud from your browser using the specialized tools we provide.

Browse our raw data

Browse our chainable transformations

Get started

>>> from kelvin import k
>>> k("WMT", "revenue", "3Q2013")
115688000000.00
>>> from kelvin import features, label
>>> from datetime import datetime
>>> as_of = datetime(2016, 9, 26)
>>> vector = features("WMT", as_of)
>>> vector[0:10]
[4, 3, 4, 6, 7, 2, 0, 1, 1, 9]
>>> label(0)
'total return trailing 1 second'
>>> label(6163)
'peer relative term frequency trailing 5 years: "underfunded pension plan"'
Search
Download
Screen
Backtest
Which type of data series do you want to download?

quantities

search terms

i
p
Configuration

firms

documents

* custom configuration

Firms

return

reset

All Public Companies

Exclude
Include
Include
Exclude

Documents

return

reset

Annual Reports

Quarterly Reports

Prospectuses

Misc Filings

Press Releases

Transcripts

News Articles

Web Pages

Social Media Authored

Social Response

Social Media About

Patent Filings

Advertisements

Consumer Reviews

Analyst Research

Include data derived from the sources selected above.

Log In

Return

Sign Up

HEADING

return

CONTENT