	In Browser
	StumbleUpon
	del.icio.us
	Google
	Google Buzz
	reddit
	LinkedIn

	Facebook
	Twitter
	Linkedin
	E-Mail

Generative AI > Google Gemini API > Supplier Risk Scoring Agent

Supplier Risk Scoring Agent

Author: Venkata Sudhakar

Supply chain resilience depends on knowing which suppliers are likely to cause disruptions. Supplier risk scoring helps procurement teams proactively manage vendor relationships before a late delivery or quality failure impacts production.

A Supplier Risk Scoring Agent uses Gemini AI to analyse each supplier across multiple dimensions - delivery performance, quality reject rates, payment reliability, and geographic risk - and produces a composite risk score with specific mitigation recommendations.

The below example shows how ShopMax India evaluates its electronics component suppliers to identify high-risk partners and prioritise mitigation actions.

import asyncio
import json
from google.adk.agents import LlmAgent
from google.adk.tools import FunctionTool
from google.adk.runners import Runner
from google.adk.sessions import InMemorySessionService
from google.genai.types import Content, Part

# ShopMax India supplier data
SUPPLIERS = [
    {
        "id": "SUP001", "name": "TechParts Mumbai",
        "on_time_delivery_pct": 94.0, "quality_reject_pct": 1.2,
        "payment_delay_days": 0, "single_source_items": 2,
        "country": "India", "years_with_shopmax": 5
    },
    {
        "id": "SUP002", "name": "ElecComp Shenzhen",
        "on_time_delivery_pct": 78.0, "quality_reject_pct": 3.8,
        "payment_delay_days": 12, "single_source_items": 7,
        "country": "China", "years_with_shopmax": 2
    },
    {
        "id": "SUP003", "name": "GlobalChip Taiwan",
        "on_time_delivery_pct": 88.0, "quality_reject_pct": 0.5,
        "payment_delay_days": 5, "single_source_items": 12,
        "country": "Taiwan", "years_with_shopmax": 8
    },
]

def score_supplier_risk(supplier_json: str) -> str:
    """Compute composite risk score for each supplier."""
    suppliers = json.loads(supplier_json)
    results = []

for s in suppliers:
        # Delivery score (0-25): lower on-time = higher risk
        delivery_score = max(0, (100 - s["on_time_delivery_pct"]) * 0.5)

# Quality score (0-25): higher reject = higher risk
        quality_score = min(25, s["quality_reject_pct"] * 5)

# Concentration score (0-25): more single-source items = higher risk
        concentration_score = min(25, s["single_source_items"] * 1.5)

# Payment / financial score (0-25): delays = higher risk
        payment_score = min(25, s["payment_delay_days"] * 1.5)

total = delivery_score + quality_score + concentration_score + payment_score

if total < 20:
            risk_level = "LOW"
        elif total < 40:
            risk_level = "MEDIUM"
        else:
            risk_level = "HIGH"

results.append({
            "supplier": s["name"],
            "risk_score": round(total, 1),
            "risk_level": risk_level,
            "delivery_risk": round(delivery_score, 1),
            "quality_risk": round(quality_score, 1),
            "concentration_risk": round(concentration_score, 1),
            "payment_risk": round(payment_score, 1),
        })

results.sort(key=lambda x: x["risk_score"], reverse=True)
    return json.dumps(results)

risk_tool = FunctionTool(func=score_supplier_risk)

AGENT_INSTRUCTION = """
You are a Supplier Risk Scoring Agent for ShopMax India procurement.
When given supplier data as JSON:
1. Call score_supplier_risk with the JSON
2. Present a ranked risk table: supplier name, total score, risk level
3. For each HIGH risk supplier, explain the main risk drivers
4. Recommend specific mitigation actions (find alternate supplier, increase safety stock, negotiate SLA)
5. Highlight any single-source concentration risks requiring urgent attention
Keep your response structured and actionable for the procurement manager.
"""

risk_agent = LlmAgent(
    name="supplier_risk_agent",
    model="gemini-2.0-flash",
    instruction=AGENT_INSTRUCTION,
    tools=[risk_tool],
)

async def main():
    session_service = InMemorySessionService()
    runner = Runner(
        agent=risk_agent,
        app_name="supplier_risk_app",
        session_service=session_service
    )
    session = await session_service.create_session(
        app_name="supplier_risk_app",
        user_id="procurement-mgr"
    )
    message = Content(parts=[Part(text="Score the risk for all our suppliers: " + json.dumps(SUPPLIERS))])
    print("Supplier Risk Scoring - ShopMax India")
    print("=" * 45)
    async for event in runner.run_async(user_id="procurement-mgr", session_id=session.id, new_message=message):
        if event.is_final_response() and event.content:
            for part in event.content.parts:
                if part.text:
                    print(part.text)

asyncio.run(main())

It gives the following output,

Supplier Risk Scoring - ShopMax India
=============================================
SUPPLIER RISK RANKING

Rank  Supplier              Score  Risk Level
----  --------------------  -----  ----------
1     ElecComp Shenzhen     48.0   HIGH
2     GlobalChip Taiwan     36.5   MEDIUM
3     TechParts Mumbai      10.6   LOW

HIGH RISK - ElecComp Shenzhen (Score: 48.0)
  Delivery risk    : 11.0 (78% on-time - well below 90% threshold)
  Quality risk     : 19.0 (3.8% reject rate is 3x acceptable level)
  Concentration    : 10.5 (7 single-source items)
  Payment risk     : 18.0 (12-day payment delay pattern)
  Actions:
  - Identify alternate supplier for top 3 single-source items immediately
  - Issue formal quality improvement notice, target < 1.5% reject rate
  - Increase safety stock by 30% for all ElecComp-supplied items

MEDIUM RISK - GlobalChip Taiwan (Score: 36.5)
  Main risk: 12 single-source items creates supply concentration risk
  Actions: Qualify a second source for at least 5 items within 90 days

CONCENTRATION ALERT
GlobalChip Taiwan supplies 12 single-source items. A Taiwan supply
disruption would impact 12 product lines. Dual-sourcing is critical.

The agent scored all three suppliers and surfaced ElecComp Shenzhen as the highest risk vendor due to poor delivery performance, high reject rates, and payment delays. GlobalChip Taiwan, despite excellent quality metrics, carries hidden concentration risk from 12 single-source items. Procurement can now prioritise mitigation actions before a supply disruption occurs.

Send your comments, suggestions or queries regarding this site to [email protected].