AI Studio Overview - Appv5 Knowledge
What's in this article?
AI Studio is the command center for building, testing, and deploying AI agents both internally and to customers. It is composed of several core modules, each designed to give your organization precise control over how AI agents are configured, branded, governed, and monitored.
Home
The AI Studio Home screen provides you with a snapshot of your AI usages within CXME. It provides filters for date ranges and shows you total usage, token usage, and cost as well as a breakdown of usage by each agent in your organization. Learn more about the Home screen in the Inbox Home Screen Overview article.
Agents
The Agents module shows you a list of all Agents within your organization. It is also where you can create and edit your agents. There are four distinct agent types, each with a dedicated no-code configuration screen.
| Agent Type | Description |
|---|---|
| Summarizer | Summarizes conversations, cases, customers, locations, and contact records with a single click. |
| Copilot | An internal agent that supports your team in finding the best answers quickly and taking actions on their behalf. |
| Conversational | Chat and voice agents that can be deployed to customer-facing screens to help automate and intelligently route conversations with customers. |
| Article Writer | Writes and updates knowledge articles on your behalf. |
You can learn more about the various agent types in the AI Agent Types in AI Studio article and how to configure agents in the Configuring Copilot and Conversational Agents in AI Studio article.
Workflows
Workflows are multi-step, deterministic agent processes that can be triggered within conversations or via events. They allow you to define structured sequences of actions that agents carry out automatically based on the conditions you set. Learn more about Workflows in the AI Studio Workflows Overview article.
Tools
Tools are API calls that agents can make to look up information or take action and make updates on your behalf. They extend the capabilities of your agents by connecting them to external systems and data sources. Learn more about tools in the AI Studio Agent Tools: Overview, Usage, and Approval Process article.
Branding
The Branding module provides a single place to define your organization's brand voice, ensuring a consistent tone across all agents and channels. Learn more about Branding in the AI Studio Branding Overview article.
Models
The Models module allows you to choose from over 340 AI models and approve only the ones you want to use. Once approved, models can be selected individually for each agent. This module also provides insight into your model usage, total spend, predicted monthly spend, and the ability to set monthly spending limits. Learn more about model selection and management in the AI Studio Models Section and Oversight article.
Observability
Observability provides robust and detailed trace data for every AI agent interaction. You can see exactly what was asked, how the agent reasoned, and how it responded to each inquiry. Learn more about Observability in the job AI Studio Observability Overview article.
Evaluations
The Evaluations module allows you to build tests for AI agents that you can run before deploying changes, ensuring continued accuracy with each update. You can define data sets of questions and expected answers, and use LLMs-as-a-Judge to verify that your agents are responding as expected after every change. Learn more about evaluations in the AI Studio Agent Evaluations Overview article.