DDictionaryLeapBuild a dictionary

From Logical Leap

A governance-ready data dictionary in minutes. No catalog. No sales call.

Upload a spreadsheet or paste a SQL schema. DictionaryLeap drafts plain-English definitions, a business glossary, a sensitivity register, suggested owner roles, and the questions your data owners still need to answer.

Build a dictionarySee a fictional example

Structure-only by default. Column names, types, and aggregate stats leave the browser. Sample values stay put unless you opt in. Profiles and packs are deleted after 24 hours.

What you get
A pack you can hand to a data owner, not a paragraph in a chat window.
  • Plain-English business definition for each column
  • Observed type, format, and example pattern
  • Suggested owner role, not a named person
  • Sensitivity and potential PII, labeled as potential, not legal advice
  • A confidence score, and a questions list where the draft is weak
  • Excel, Word, Markdown, and dbt schema.yml on the paid pack
Fictional sample

Before and after

Harbor & Pine Outfitters is not a real company. The rows below are a designed example of the pack, not a live generation from your data. The last row shows a name the draft refuses to invent a meaning for.

Before
A header row from a commerce extract.
order_id,cust_email,ship_zip,ord_dt,qty,flg_03
10482,ada@harborpine.example,98101,2026-02-11,2,Y
After
The same columns, ready for an owner to review.
ColumnExampleDefinitionTypePatternOwner roleSensitivityPIIConfidence
order_id10482Order identifierInteger9 (sample)Order operations leadInternalNo88%
cust_emailada@harborpine.exampleEmail address for the customerTextemailCustomer data stewardConfidentialPotential email90%
ship_zip98101Postal code for the shipmentText99999Order operations leadInternalPotential postal code84%
ord_dt2026-02-11Date of the orderDatedateOrder operations leadInternalNo86%
qty2Count of units on the lineInteger9Order operations leadInternalNo86%
flg_03YNot stated as factBooleanYSee questionsUnclassifiedNo34%

Question for the data owner: what does flg_03 mean, and which role owns it?

1. Add a file or a schema

CSV, Excel, or a SQL CREATE TABLE script. Profiling runs in the browser.

2. Read the draft with the confidence on it

Strong names are stated. Weak names become questions instead of facts.

3. Keep the preview, or export the pack

The first 25 columns stay on screen. $49 unlocks the full file, up to 200 columns.

Straightforward pricing

Start with the preview. Pay only if you want the exports. Files are deleted after 24 hours either way.

Free preview
One file, first 25 columns, on screen.

$0

  • Watermarked results
  • Glossary, sensitivity, and questions for those columns
  • Exports turned off
Try the preview
Dataset Pack
One file, up to 200 columns
The full governance pack, ready to download.

$49

  • Excel, Word, Markdown, and dbt schema.yml
  • Dictionary, glossary, sensitivity register, questions
  • Emailed download link, deleted after 24 hours
Preview, then unlock exports
Coming soon
Database Pack
Up to 75 tables or 2,000 columns.

$249

  • Schema exports and multi-table files
  • Same pack, across the database
  • Not on sale yet

FAQ

Structure-only mode is the default. DictionaryLeap profiles the file in your browser and sends column names, inferred types, and aggregate stats such as null percent, distinct count, and pattern shapes. Sample values stay on your machine unless you turn on the optional toggle. Those samples are redacted before they are sent. Uploaded names and values are treated as data, not as instructions. Profiles and generated packs are deleted after 24 hours. DictionaryLeap does not train models on your files.