Call Center Youth Summary Generator¶

This workflow provides call center agents with comprehensive youth entrepreneur profiles by extracting phone numbers from queries and generating detailed summaries that include personal information, business performance, credit status, active alerts, and recent conversation history.

Purpose¶

No business context provided yet — add a context.md to enrich this documentation.

This workflow serves call center agents who need quick access to complete youth entrepreneur profiles. When an agent receives a query containing a phone number, this workflow automatically retrieves and formats all relevant information about that youth, including their business performance trends, credit management status, any pending alerts, and recent AI conversation history. This enables agents to provide informed, personalized support without manually searching through multiple systems.

How It Works¶

Query Processing: The workflow receives a text query from another workflow and uses AI to extract any phone number mentioned
Phone Normalization: Converts the extracted phone number to standard Kenyan format (+254XXXXXXXXX)
Validation: Checks if a valid phone number was found and provided
Profile Lookup: Searches the youth entrepreneurs database for the normalized phone number
Data Aggregation: If found, retrieves related data from multiple tables:
- Youth profile information (personal details, business info, current status)
- Recent profit tracking data (last 14 days of business performance)
- Credit management status and micro-actions
- Active alerts requiring attention
- Recent AI chat history
Summary Generation: Formats all collected data into a comprehensive, readable summary
Response: Returns either the complete summary or appropriate error messages

Workflow Diagram¶

graph TD
    A[When Executed by Another Workflow] --> B[Information Extractor]
    B --> C[Normalize Phone]
    C --> D[If Phone Valid]
    D -->|Valid| E[Query Youth Profile]
    D -->|Invalid| F[Return No Phone]
    E --> G[If User Found]
    G -->|Found| H[Query Profit Data]
    G -->|Not Found| I[Return Not Found]
    H --> J[Query Credit Status]
    J --> K[Query Active Alerts]
    K --> L[Query Recent Chat]
    L --> M[Format Summary]
    M --> N[Return Summary]
    F --> N
    I --> N

    O[OpenRouter Chat Model] -.->|AI Language Model| B

Trigger¶

Execute Workflow Trigger: This workflow is designed to be called by other workflows, receiving input data through the passthrough mode.

Nodes Used¶

Node Type	Purpose
Execute Workflow Trigger	Receives input from calling workflows
Information Extractor (LangChain)	Uses AI to extract phone numbers from text queries
OpenRouter Chat Model	Provides AI language model for information extraction
Code (Normalize Phone)	Converts phone numbers to standard Kenyan format
If (Phone Valid)	Validates extracted phone number
If (User Found)	Checks if youth profile exists in database
Postgres (Multiple)	Queries various database tables for youth data
Code (Format Summary)	Processes and formats all data into readable summary
Set (Multiple)	Returns appropriate responses for different scenarios

External Services & Credentials Required¶

OpenRouter API¶

Credential Name: call_center_get_summary (OpenRouter API)
Purpose: Powers the AI information extraction
Required: API key for OpenRouter service

PostgreSQL Database¶

Credential Name: Postgres account
Purpose: Access to youth entrepreneur database
Required: Database connection details (host, port, database name, username, password)

Environment Variables¶

No specific environment variables are configured in this workflow. All configuration is handled through n8n credentials.

Data Flow¶

Input¶

Format: JSON object with query field
Example: { "query": "Please get summary for youth with phone 0712345678" }

Output¶

Format: JSON object with summary field
Success Example: Detailed multi-section summary including youth profile, status, profit trends, credit status, alerts, and chat history
Error Examples:
- "No phone number was provided. Please provide the youth's phone number to retrieve their summary."
- "No youth found with phone number +254712345678. Please verify the number and try again."

Error Handling¶

The workflow includes comprehensive error handling:

Invalid Phone Numbers: Returns helpful message when no valid phone number is extracted
User Not Found: Provides clear feedback when phone number doesn't match any youth profile
Database Errors: All database queries use alwaysOutputData and continueRegularOutput to handle connection issues gracefully
Missing Data: The formatting logic handles null/empty values with fallbacks to "N/A"

Known Limitations¶

Phone number extraction relies on AI and may occasionally miss numbers in complex text
Database queries are limited to recent data (14 days for profit tracking, 10 messages for chat history)
Summary formatting is optimized for English text and Kenyan phone number formats
No caching mechanism - each request queries the database fresh

This workflow is designed to be called by other workflows in the call center system. The calling workflow should pass queries containing phone numbers through the trigger.

Setup Instructions¶

Import Workflow: Import the JSON into your n8n instance
Configure Credentials:
- Set up OpenRouter API credential named call_center_get_summary
- Configure PostgreSQL credential named Postgres account with access to:
  - youthEntrepreneursReal table
  - dailyProfitTracking table
  - creditTracking table
  - ceaAlerts table
  - n8n_chat_histories table
Database Requirements: Ensure your PostgreSQL database contains the required tables with the expected schema
Test the Workflow:
- Execute manually with test data: { "query": "Get summary for 0712345678" }
- Verify database connections work
- Check that phone number normalization works for your region
Integration: Call this workflow from other workflows using the Execute Workflow node, passing queries in the expected format
Monitoring: Monitor execution logs to ensure AI extraction and database queries perform as expected