PROFILE → CONVERSE → FIX

Your data has
opinions about
your pipeline.

Upload a CSV or a document collection. Get a schema, a ranked issue list, and fix rules you can read, edit, and reverse, before any of it reaches your warehouse or your RAG index.

no credit card · early access, rolling invites

clients_export_q3.csv5 columns flagged
signup_emailstring14% malformedtrim + regex_replacecritical
plan_tiercategorical"N/A", "--", "" mixednullify_placeholdersimportant
renewal_datedate3 formats detectedcoerce_dateimportant
account_ownerstringcase inconsistentcase: titleminor
zip_codestringleading zeros droppedtrim + padminor
5 rules queued · nothing applied until you approve it

We ran a competitor’s cleanup tool against a real export. Every single null_handling rule it wrote failed, silently, with the same error:

Unknown null_handling op: 'nullify_placeholders'

The operation was referenced in the rule set. It was never implemented.

WHY THIS EXISTS

Data Readiness implements nullify_placeholders, trim, case, and regex_replace for real, and shows you the exact rule, in order, before it touches a single row.

01

Structured

Upload a CSV. We infer schema and semantic types, surface issues ranked by severity, and propose a transform rule per column.

  1. infer schema + semantic types
  2. review issues & rules per column
  3. apply fixes
  4. download cleaned CSV + fix summary + workflow doc

02

Unstructured

Upload documents: PDF, DOCX, TXT, EML, MSG, HTML. We profile the collection for RAG-suitability and hand you a prioritized fix list.

  • duplicates
  • chunking quality
  • missing metadata
  • staleness
rulesinspectable & reversible, remove any rule before it runs
modeltiered by task: cheap model for per-value fixes, stronger model reserved for schema judgement
severitycritical / important / minor, always labeled, never color alone
outputcleaned file + fix summary + workflow context doc, every run
all posts →

Clean your data.
Trust your forecasts.