Supplier Scorecard Builder
Turn scattered supplier quotes and email threads into defensible, weighted scorecards in minutes.
Which market this reads
This check reads no market's rules, so no jurisdiction applies to what it returns.
Executes user-defined scoring and landed cost normalization over user-provided supplier quote data; reads no statutory register or national regulation.
What it does
Evaluates FMCG supplier quotes and performance records, normalises landed costs across currencies and incoterms, and generates weighted scorecards.
Ask it like this
youBuild a supplier scorecard comparing Alpha Pack (1.20 EUR DDP, 14 days lead time) and Beta Fill (1.05 EUR EXW, 28 days lead time).
You never name it yourself. Your assistant picks the right tool: supplier_scorecard_builder__build_supplier_scorecard
What you give it
| What to give | Kind of value | Needed? | Notes |
|---|---|---|---|
| quotes | array | yes | supplier quote records with unit_price, currency, incoterm, lead_time_days, delivery_reliability_pct, payment_terms_days |
| base_currency | string | optional | target base currency code for landed cost comparison (default EUR) |
What you get back
| What comes back | Kind of value | Notes |
|---|---|---|
| ranked_suppliers | array | suppliers ordered by overall weighted score with landed cost |
| confidence | number | 0–100 |
| review_required | boolean | true when confidence is below threshold or required fields missing |
| disclaimer | string | states user rubric was applied, not legal compliance or certification |
Sources
- ISO 9001:2015 Clause 8.4 control of externally provided processes, products and services
- VDA 6.3 quality standard for process audit and supplier evaluation
Who wrote this
Every capability is written by one named publisher. You see who stands behind an answer, and you can take the conversation further with them directly.
fmcg.network Core Team
verifiedPlatform capabilities and reference implementations
Network team
What the assistant can offer next
Related reading
How uncertainty is handled
This check reports a confidence score with every answer. Below 80 out of 100 nothing is guessed: the uncertain items go to a review queue and the answer comes back marked "needs review".
Read how quality levels work