Customer Support Agent

Viewing the same solution through different personas

Navigation

🏠 Go to Home ⌘H

🛒 Open Flexi Store ⌘S

Actions

💾 Save Solution ⌘S

🚀 Deploy Solution

📋 Duplicate Solution

Persona Switching

💻 Switch to Pro Code ⌘1

🎨 Switch to Low Code ⌘2

✨ Switch to No Code ⌘3

View Mode: Switch personas to see different abstractions of the same solution

Flexi

Customer Support Agent

Online • Ready to help

Flexi

Hello! I'm your Customer Support Agent. I can help you with inquiries, technical issues, billing questions, and more. How can I assist you today?

I need help with my order

Flexi

I'd be happy to help with your order! Could you please provide your order number? I can look up the details and assist you further.

Agent Configuration

Name: Customer Support Agent

Version: 2.1.0

Model: GPT-4

Temperature: 0.7

Memory: 24h retention, 4000 token context

Escalation: Enabled for complex issues

Workflow Execution

Input Received

User query processed

LLM Processing

GPT-4 analyzing intent...

Routing

Route by intent

Memory Store

Save context

Response

Format output

Explorer

src

system_prompt.yaml

agent_config.json

workflows

system_prompt.yaml

agent_config.json

1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950

system:
  name: "Customer Support Agent"
  version: 2.1.0
  description: "Professional customer support agent with escalation and context management"
  
persona:
  tone: "professional, empathetic, solution-oriented"
  expertise: "customer service, problem-solving, technical support"
  communication_style: "clear, concise, helpful"
  
capabilities:
  - "handle customer inquiries"
  - "analyze issue severity"
  - "escalate complex issues to human agents"
  - "maintain conversation context"
  - "provide product information"
  - "process refunds and returns"
  
workflow:
  entry_point: "user_query"
  steps:
    - step: 1
      action: "receive_input"
      type: "input"
    - step: 2
      action: "llm_process"
      type: "llm"
      model: "gpt-4"
      temperature: 0.7
    - step: 3
      action: "route_by_intent"
      type: "router"
      routes:
        - "general_inquiry"
        - "technical_issue"
        - "billing_question"
        - "escalation_required"
    - step: 4
      action: "store_context"
      type: "memory"
      retention: "24h"
  
memory:
  type: "conversational"
  retention: "24h"
  context_window: 4000
  
escalation:
  enabled: true
  triggers:
    - "complex_technical_issue"
    - "billing_dispute"
    - "customer_request"
  escalation_target: "human_agent_queue"
  
api_integrations:
  - name: "product_database"
    endpoint: "https://api.example.com/products"
  - name: "order_system"
    endpoint: "https://api.example.com/orders"
  
response_format:
  style: "structured"
  include_suggestions: true
  max_length: 500

Terminal

$ flexigen validate system_prompt.yaml

✓ Configuration valid

$ _

Components

LLM

LLM Node

⤢

Splitter

⇄

Router

💾

Memory

API

API Call

⚙

Transform

Customer Support Agent Workflow Visual representation of the same solution

Step 1: Input

User Query
receive_input

Step 2: LLM Process

GPT-4
Temp: 0.7
llm_process

Step 3: Intent Router

Route by Intent
route_by_intent

Step 4: Memory

Store Context
24h Retention
store_context

Step 5: Response

Formatted Output
Max 500 chars

General Inquiry

Handle directly

Technical Issue

Analyze & resolve

Billing Question

Process request

Escalation

Human Agent Queue
Complex issues

Selected Node

Model

Temperature 0.7

Max Tokens

System Prompt

Professional customer support agent with empathetic tone. Analyze user queries and route appropriately.

Context Window

Customer Support Agent - No Code View

This is the same solution shown in a simplified, guided interface

Solution Configuration

Solution Name: Customer Support Agent

Version: 2.1.0

Solution Type: Customer Support & Service

Persona: Professional, Empathetic, Solution-Oriented

Core Capabilities: Handle Inquiries • Analyze Issues • Escalate Complex Cases • Maintain Context

Workflow Steps: Input → LLM Process → Route by Intent → Store Context → Response

LLM Model: GPT-4 (Temperature: 0.7)

Memory: 24-hour Conversational Retention (4000 token context)

Escalation: Enabled (Complex Issues, Billing Disputes, Customer Requests)

API Integrations: Product Database • Order System

Router Paths: General Inquiry • Technical Issue • Billing Question • Escalation Required

Same Solution, Different View

Switch to Pro Code or Low Code to see the technical details of this same Customer Support Agent solution.

Customer Support Agent

Preview: Customer Support Agent

Customer Support Agent

Agent Configuration

Workflow Execution

Customer Support Agent - No Code View