Getting Started with ResponseCX
An introductory guide to building powerful AI agents with ResponseCX.
What is ResponseCX?
ResponseCX is a platform for building and deploying autonomous AI agents, especially for e-commerce businesses using platforms like Shopify. These agents can handle customer service inquiries, answer questions by accessing knowledge bases and APIs, and provide 24/7 support to your customers.
✨ Key Features
- Conversational AI Agents – Configurable GPT-4/GPT-4o agents with tool-calling (Reasoning API) and memory (mem0) support.
- Omni-Channel Routing – Unified inbox for web-chat, email, SMS, WhatsApp and more.
- Knowledge Fusion – Pulls context from knowledge-bases, GraphQL APIs and vector stores to ground responses.
- Composable Workflows – Trigger TypeScript/GraphQL actions, webhooks or third-party APIs from any conversation event.
- Kubernetes-Native – First-class Helm charts, horizontal scaling and stateless architecture.
- Open & Extensible – Built with Next.js, TailwindCSS and GraphQL; override or extend any component.
🏗️ Architecture Overview
- Next.js App – UI & API Routes served via Vercel or Kubernetes pod.
- Reasoning API – Enables tool-calling and agent reasoning chains.
- Providers – GraphQL, REST or vector search abstractions.
- Real-time Transport – Server-Sent Events (SSE) stream agent output to clients.
How It Works
At its core, ResponseCX follows a simple but powerful process:
- Interpret: It understands the customer’s question, comment, or request.
- Generate: It formulates the best response by consulting your company’s knowledge, which can include FAQs, documents, and even external APIs.
- Respond: It delivers the correct answer to the customer in real-time.
Your First Steps in ResponseCX
Ready to build your first agent? Follow these steps to get started.
Sign In & Create an Organization
First, sign in to ResponseCX. Once you’re in, create an Organization. Your current organization is displayed in the bottom-left corner.
Create a New Agent
Navigate to the Agents tab and click the Add New Agent button. Give your agent a name and a purpose. For more details, see our Agents Quickstart.
Start a Chat
Click the chat icon in the top-left corner to start a new conversation. Give your chat a name, and select the agent you just created. You can also choose the underlying language model. We support models like GPT-4o, Llama3-70b, and Claude 3.5 Sonnet.
Building Your Agent’s Knowledge
An agent is only as smart as the information it has access to. You can build your agent’s knowledge base directly from the chat interface.
!add
: Adds a new piece of information.!update
: Modifies existing information.!delete
: Removes a piece of information.
Example:
!add The return policy is 30 days for a full refund.
After adding data, you can immediately test the agent by asking questions about it.
For more advanced knowledge management, you can create Sources. Learn more in our Sources Quickstart.
Customizing Agent Behavior
You can fine-tune your agent’s personality and decision-making using Attributes and Rules.
Agent Attributes
Attributes define your agent’s characteristics.
Select Your Agent
Go to the Agents tab and select the agent you want to configure.
Define the Attribute
Click on Attributes and then Add Attribute. Define its type, name, and description.
Set the Attribute Value
Use the slider or input field to assign a value to the attribute.
Save the Attribute
Click Add Attribute to save it.
Agent Rules
Rules are conditional statements that guide your agent’s behavior.
Select Your Agent
Go to the Agents tab and select the agent.
Create the Rule
Click on Rules and then Add Rule. Give the rule a name, a description, and define the condition that triggers it.
Save the Rule
Click Add Rule to save.
Next Steps
You’ve now learned the basics of creating and customizing an agent in ResponseCX. Here’s what you can do next:
- Dive deeper into agent configuration.
- Learn about connecting different information sources.
- Explore how to create complex rules and schedules for your agents.