Skip to content
Data dump to structure / Sort: Automatic classification
Open app

Sort: Automatic classification

The moment a file enters the pipeline, Colabra classifies it into a specific category and subcategory. Classification is automatic and gives the team a consistent way to label, scan, and filter files in the evidence list.

Each category has its own icon and colour in the app, so you can scan a file list at a glance.

Evidence files list

Top level category list

CategoryWhat it covers
CorporateFormation docs, bylaws, minutes, org charts
ContractsAgreements, amendments, schedules, side letters
FinancialStatements, management reports, projections
TaxReturns, filings, opinions, transfer pricing
HREmployment agreements, policies, headcount
IPPatents, trademarks, licences, trade secrets
ITSystems docs, security policies, infrastructure
RegulatoryPermits, licences, regulatory filings
EHSEnvironmental, health, and safety
CommercialCustomer and supplier agreements, revenue
OperationsProcedures, supply chain documentation
Real estateLeases, deeds, surveys, property records
InsurancePolicies, claims history, certificates
LitigationPleadings, settlements, legal opinions
CommunicationsEmails, memos, internal correspondence
TransactionDeal documents, LOIs, term sheets
SeparationTSAs, separation plans
MiscellaneousDocs that don't fit a specific category

Sorting is what makes the first pass usable

Example workflow: reviewers start from lanes, not from a flat file dump

After a large upload, legal reviewers want to look at contracts, finance wants statements and schedules, and compliance wants regulatory material. Automatic classification is what turns one incoming batch into reviewable lanes instead of one undifferentiated evidence list.

File formats and limits

The file-format matrix and the practical upload and preview limits now live in File formats and limits.

That page covers:

  • which formats are parsed vs. previewed
  • which formats are only recognized or display-only
  • upload, parsing, and preview size limits
  • spreadsheet row, sheet, and cell safety limits

Reviewing and correcting classification

Review classification when it affects your queue. Edge cases occur — a document labelled “miscellaneous” might actually be a key commercial agreement, or a financial report might be categorised as corporate.

Check the classification assigned to each file from the Evidence tab. If the AI miscategorised a file, correct it so the evidence list reflects the right label for the team. Changing a category is an organizational edit; it does not by itself establish that every required structured extraction has run. Open the file’s available analysis tabs and check processing separately. Changing a category is an organizational edit; it does not by itself establish that every required structured extraction has run. Open the file’s available analysis tabs and check processing separately.