What this category's data looks like
Laboratory hazardous chemicals procurement and ledger materials cover three types of forms: procurement applications, usage registrations, and inventory ledgers. Procurement applications record name, specification, quantity, application date, and approval signature; usage registrations record the user, quantity used, date of use, and custodian signature; inventory ledgers record the balance quantity, update date, and inventory count signature. Materials are updated by batch and version, and the applications, usage registrations, and ledgers for the same batch often belong to different files. When entering this function, the three types of materials are imported as independent data rows, each generating vectors and retaining the original data index fields, to support subsequent retrieval and citation trimming for supporting materials.
Configuration and implementation checks
This step is executed by the indexing service and vector database, outputting vectors and index status (whether recall is possible, whether data in the same group is merged). The basis entries that can be traced back to the original text are then completed by subsequent AI nodes, code nodes, or manual work (retrieval and citation trimming only take supporting materials).
| Object | Check | Verification action |
|---|---|---|
| Procurement application materials | Name, quantity, date, and signature can be reconciled with the ledger; the version is current | Check whether the name and quantity on the application form match the ledger, check whether the approval signature and receipt are complete, and confirm that the version date is current |
| Usage registration materials | The user, quantity used, date, and custodian signature can be reconciled with the inventory ledger | Check whether the quantity and date in the usage record match the changes in the ledger balance, and check whether the custodian signature and usage receipt are complete |
| Inventory ledger materials | The balance quantity, update date, and inventory count signature can be reconciled with the application and usage records | Check whether the ledger balance is consistent with the application and usage records, and check the inventory count signature and version update records |
| Data row index fields | Multiple vector identifiers for one data row are all present, and during retrieval data in the same group is merged and the highest score is taken | Open a data row and check the index fields, run a retrieval once to see whether data in the same group is merged; check the status value indexed and the rebuild task records, and check limit, and the tokens of the cited text |
Three common errors
- Symptom: The name, quantity, and date of the same record do not match across two materials. Inspect the following: Procurement applications, usage registrations, and inventory ledgers belong to different files, and reconciliation checks were not performed.
- Symptom: Versions are not synchronized, and old-version materials are still recalled during retrieval. Inspect the following: After changing the model, the rebuild task is triggered only for data whose indexing has already been completed, and does not cover all data rows.
- Symptom: Signatures or receipts are missing, and basis entries cannot be traced back to the original text. Inspect the following: Approval signatures, custodian signatures, or usage receipts are missing, and the materials are incomplete.
How to verify the setup
This function's acceptance check: take one data row and look at the index fields, check whether multiple vector identifiers are all present, then run a retrieval once to see whether data in the same group is merged and the highest score is taken; check the status value indexed and the rebuild task records, and check limit, and the tokens of the cited text.
Business test: ask a basis question that spans materials, see whether the answer gives a traceable source, and check the number of citations and the tokens of the cited text; record the array length separately.
- Check whether the indexing model is consistent with the vector model actually used by the knowledge base, and view this in the knowledge base settings.
- Open a data row and check the index fields to confirm that multiple vector identifiers are all present; check the data scope covered by the rebuild task.