Data
Encoder

Transcribes from the messy originals — scanned supplier invoices, handwritten job dockets, paper application forms, PDFs that won't copy — into the client's system, keying to the exact field format rather than a close approximation.

Finds duplicate contacts, companies and products using match rules on email, phone, ABN and address, applies the client's survivorship rules for which values survive the merge, and merges in batches with a record of what was joined.

Standardises records to Australian addressing — unit and street number, street type, suburb, state and four-digit postcode — validates the suburb and postcode pairing, and normalises mobile and landline numbers to one consistent format.

Checks each business record against ABN Lookup for the registered entity name, ABN status, GST registration and recorded trading names, and corrects the records where a trading name has been entered as the legal entity.

Prepares CSV import files against the system's template, cleans and reshapes the data in Excel or Power Query, tests on a small sample first, then loads the full file and reconciles the row count before and after.

Builds the field-by-field mapping from the old system to the new one, stages the data, works out where two old fields collapse into one, and reconciles totals so the client can sign off before the old system is switched off.

Loads and maintains SKU records — code, description, unit of measure, cost and sell price, supplier, category, images and specification attributes — into the inventory or e-commerce platform, keeping naming consistent across the whole catalogue.

Indexes scanned and OCR'd documents with the metadata the business actually searches on — client, job number, document type, date — and files them into SharePoint or the document management system under the agreed naming convention.

Handles records that fail validation — missing mandatory fields, invalid dates, amounts outside the expected range, unmatched references — researches the correct value from the source document, and refers back to the record owner where it can't be determined.

Runs a standing set of checks for blank mandatory fields, orphaned records, contacts with no owner and impossible dates, then reports the exceptions each week with what was corrected and what needs a decision from the client.

Keeps the data dictionary current: what each field means, the picklist values allowed, the format expected and who owns it, so the same information is entered the same way by everyone touching the system.

Balances each keyed batch back to the source totals — invoice count and value, timesheet hours, transaction sums — before marking it complete, so an error is caught at the batch rather than in a month-end report.

Pulls the scheduled data extracts the business relies on, applies the agreed filters and date ranges, formats them the same way every time, and delivers them to the same place on the same day each week or month.