Skip to main content
Procurement Quote Comparison is a community Plugin App for AI-assisted procurement comparison. It supports enterprise purchasing, administrative purchasing, IT procurement, and finance shared-service scenarios. Users create a comparison case in Xpert Project or Agent Workbench, let an Assistant parse requirement and supplier quote documents, and review structured results, risks, and recommendations.

When To Use It

  • Procurement users need to compare multiple supplier quotes with inconsistent file formats.
  • Requirement documents contain items, specifications, quantity, budget, and expected delivery dates.
  • Supplier quotes need structured supplier, contact, tax, delivery, payment, warranty, and line-item data.
  • Users need to detect specification mismatch, quantity mismatch, missing quotes, delivery risk, warranty risk, and price anomalies.
  • Users need explainable answers to questions such as “why not choose the lowest price?”

Plugin URL

Marketplace: Procurement Quote Comparison

What The App Adds

TypeNamePurpose
Project / Agent Workbench viewProcurement Quote Comparison WorkbenchManage comparison cases, upload documents, trigger parsing, and review comparison results and AI recommendations.
Assistant templateProcurement Quote Comparison AssistantParse requirements and supplier quotes, generate item comparisons, risk findings, and recommendation reports.
Assistant toolsProcurement Quote Comparison ToolsSave requirement parsing, quote parsing, item matches, risk items, final recommendations, and parse failures.

System Architecture

Procurement Quote Comparison uses a comparison case as the business boundary. Workbench stores requirement and supplier quote file handles, the Assistant reads one file at a time and calls middleware tools, and plugin services write requirements, quotes, matches, risks, and recommendations back to the same case.

Comparison Flow

The App parses each supplier quote as a separate source document so multiple quotes do not contaminate each other in one Assistant message. The final recommendation is not just the lowest price; it also weighs specification fit, quantity, tax, delivery, warranty, payment terms, and risk findings.

1. Create a comparison case

Open the procurement comparison entry in Xpert Project or Agent Workbench and upload a requirement document. The App creates a comparison case from the file and stores the platform file handle so the original file can be attached to Assistant parsing messages. Manual case creation is also available, but starting from the requirement document is the recommended path.

2. Upload supplier quote files

Open the case detail and upload quotes from at least two suppliers. Each quote is saved as a separate source document. During batch parsing, each Assistant message attaches only one supplier quote file, preventing data from different suppliers from contaminating one another.

3. Parse requirements and quotes

Select an Xpert with procurement comparison middleware and trigger:
  1. Requirement parsing.
  2. Supplier quote batch parsing.
  3. Comparison result generation.
The Assistant calls procurement_save_requirement to save requirement fields and item rows, then procurement_save_supplier_quote to save supplier quote headers, terms, and line items.

4. Generate comparison and recommendation

After parsing, the Assistant matches supplier quote items to requirement items, then saves item matches, risks, and final recommendations. The workbench shows requirements, supplier quotes, comparison tables, risk findings, recommended supplier, recommendation rationale, report draft, and pending questions. When AI fields conflict with manually maintained fields, the App preserves manual values and records the conflict as a pending difference.

Tool Boundaries

ToolPurpose
procurement_save_requirementSave procurement project fields and requirement line items.
procurement_save_supplier_quoteSave supplier quote headers, commercial terms, and quote line items.
procurement_save_item_matchesSave quote-to-requirement item matching results.
procurement_save_risk_itemsSave specification, quantity, price, delivery, warranty, and other risk findings.
procurement_finalize_recommendationSave the final AI recommendation, explanation, report draft, and pending questions.
procurement_report_parse_failureRecord parse failures for later retry.

Data Objects

ObjectMeaning
ProcurementComparisonCaseBusiness comparison case and isolation boundary.
ProcurementSourceDocumentRequirement and supplier quote source files.
ProcurementParseJobParsing task and status.
ProcurementRequirementItemRequirement line item.
ProcurementSupplierQuoteSupplier quote header and commercial terms.
ProcurementQuoteItemSupplier quote line item.
ProcurementItemMatchMatch between requirement item and quote item.
ProcurementRiskItemRisk finding.
ProcurementRecommendationRecommendation and explanation.

Best Practices

  • Upload at least two supplier quotes for meaningful comparison.
  • Parse each supplier quote separately to avoid cross-supplier contamination.
  • Keep evidence for specification, model, quantity, delivery, and warranty mismatches.
  • Lowest price should not automatically become the recommendation; delivery, warranty, specification match, payment terms, and risks matter.
  • If no procurement Xpert is configured, create the Assistant from the template and grant middleware access first.

Current Boundaries

The first version does not include ERP, SRM, OA, approval workflow, real ordering, supplier master data, historical price analysis, budget commitment, invoice verification, legal contract review, or payment workflow. It is best used to validate procurement document understanding, comparison, risk identification, and recommendation explanation.