Supplier Selection Agent
Agent Description:
The Supplier Selection Agent helps procurement teams evaluate and rank suppliers based on critical performance metrics, enabling data-driven selection of optimal suppliers. It automates the scoring and ranking process using a weighted model, then recommends the top supplier(s) along with a clear justification for the selection.
- Purpose: To identify and recommend the best-fit supplier(s) based on a weighted evaluation of cost, quality, delivery, and performance.
- Components:
- An evaluator sub-agent that applies a scoring model to supplier attributes.
- A recommender sub-agent that selects and justifies the final supplier based on ranking.
- Fetching supplier data via an API.
- Normalizing and scoring each supplier based on key attributes.
- Applying a weighted ranking formula for supplier evaluation.
- Recommending the top supplier(s) with rationale.
- Suggesting backup options if scores are close.
- Google VertexNote: To learn more about the LLM and to modify its behavior, refer to the Configuring LLM settings section.
Sub-Agents
1. Supplier Evaluator
- Role: Supplier Evaluator.
- Scope: Retrieve supplier data and calculate scores using a weighted ranking formula.
- Description:
-
Connects to a REST API using the GET Request Tool to retrieve the supplier dataset containing: Normalize all fields (if not already on a 0–100 scale).
-
Apply the weighted formula: Rank suppliers in descending order of Total Score.
-
- LLM Used: Google Vertex (inherits from parent).
- Tools used: GET Request Tool
2. Supplier Recommender
- Role:Supplier Recommender.
- Scope: Select and justify the top-ranked supplier based on performance metrics.
- Description:
-
Accepts the ranked supplier list from the Evaluator.
-
Selects the top-performing supplier based on total score.
-
Justifies the choice by referencing:
-
Total Score
-
Critical metrics: Cost, Product Quality, On-Time Delivery
-
-
- LLM Used: Google Vertex (inherits from parent).
- Tools used: GET Request Tool
- The Supplier Evaluator agent initiates the process by fetching supplier data using the GET API tool.
- It normalizes the values, applies the weighted scoring formula, and ranks suppliers in descending order.
- The Supplier Recommender agent takes this ranked list, selects the top supplier, and justifies the recommendation using score breakdowns.
- If needed, it also suggests a backup supplier if the score gap is small.
- The final output includes the selected supplier name, total score, and a concise justification summary.
- Which supplier has the best combination of quality and on-time delivery?
- What is the weighted score of each supplier using the 20/30/30/20 formula?