SPIR spare parts record extraction
SPIR forms are the least spreadsheet-like spreadsheets in engineering: multiple tabs, merged headers, tag groups running across the columns and part line items running down the rows. AxioExtract reads them with a deterministic template parser rather than a vision model, so the same form gives the same result every time.
How the read works
Specific tabs only
Matrix to triples
PDF SPIRs
Tag register and bill of materials
A tag group covers several physical tags. “Generate tag register” explodes each group into its individual tags and can be downloaded as CSV. From there the BOM panel builds either a per-tag bill of materials — every tag with its parts and quantities — or a consolidated one that sums identical parts into a single procurement line. A dropdown filter shows the item count per tag so you can view the BOM for one tag at a time. More on BOM output is on extract a BOM from a PDF.
Scoring and blank cells
SPIR forms are legitimately sparse: most tag-versus-part intersections are empty because that tag does not take that part. Completeness scoring therefore counts only the columns a SPIR read must fill, and ignores non-required cells that are blank on the source, so a correctly read sparse form is not penalised.
Previewing the workbook
Spreadsheet and PDF SPIRs are shown in a built-in preview with a tab strip and sticky headers, so a selected row scrolls into view on the original sheet. Corrections are made in the grid before exporting to Excel, Google Sheets, JSON or XML.
Common questions
What is a SPIR?
Can a SPIR be read without AI?
Can a bill of materials be produced from a SPIR?
Try it on your own document
Credits are included with every new workspace — no card required to start.
Start extracting