Cloud BOT Operator is an AI agent that automatically performs browser operations based on natural language instructions.
The AI agent runs as part of a BOT scenario and can flexibly handle everything from operations such as clicking, data entry, and copy-and-paste to complex tasks such as data processing and extraction, all in accordance with the given instructions.
Two built-in modes enable flexible browser operations.
By precisely restricting the execution scope of the AI agent, you can prevent unintended operations and safely automate only the necessary areas.
You can configure the range of operations permitted for the AI agent in three levels.
| Setting | Description | Use Case |
|---|---|---|
| Full Access | Freely operates the browser according to the prompt | Form input, data registration |
| Read-Only Access | Browser operations prohibited. Screen reading only | Screen data extraction, status monitoring |
| No Access | Browser access prohibited. Data processing only | Text data processing |
Restricts the screen area accessible to the AI agent. By limiting viewing and operations to within the specified range, it prevents accidents caused by misrecognition while also reducing processing costs.
Enabling this mode for cases that require more complex situational judgment improves the AI agent's decision-making accuracy. It is effective for irregular screen layouts that are difficult to handle in standard mode, and for operations involving multiple conditions. Cost increases when enabled.
The AI automatically identifies operations that can be processed together—such as form input and data copying—and executes them in a single step. By reducing the number of processing steps, costs are lowered, making it especially effective for tasks with many repetitive operations.
Allows the AI agent to generate and execute JavaScript as needed. This enables advanced processing such as bulk data updates and DOM manipulation that would be difficult with standard browser operations, greatly expanding the scope of automation.
You can select the optimal AI model based on your use case and cost requirements.
| Model Name | Features | Speed | Cost | Selectable AI Models |
|---|---|---|---|---|
| Simple | The lightest model for simple operations. | Fastest | Low |
・gpt-5.4-nano ・gemini-3.1-flash-lite-preview |
| Assist | A model for standard operations. | High-speed | Standard |
・gpt-5.4-mini ・gemini-3-flash-preview |
| Control | A high-performance model for complex operations. | Low-speed | High |
・gpt-5.4 ・gemini-3.1-pro-preview |
| Expert | A model for the most advanced and difficult operations. | Low-speed | High | ・gpt-5.5 |
Reads data from CSV files and automatically enters it into the target website. The AI agent automatically identifies input fields and handles data correction and conversion, enabling flexible handling even when form structures are complex or field names do not match.
AI's role: Identifying input fields on the website, data correction and conversion
Extracts and summarizes necessary information from received emails and automatically transfers it to the target website. Even when email content varies each time, the AI agent interprets the content and enters it into the appropriate fields.
AI's role: Data extraction and summarization from email body, identifying input fields on the website
Automatically logs into a website and periodically checks the screen status. The AI agent evaluates the content and sends an email notification if an issue is detected. This automates manual visual checks and prevents oversights.
AI's role: Reading screen status and evaluating content, anomaly detection
Cloud BOT realizes hybrid browser automation combining RPA and AI agents. The AI agent handles flexible judgment and instruction interpretation, while RPA handles stable and low-cost routine operations, enabling automation that maximizes the strengths of each.
By combining RPA and AI agents according to your use case and business needs, you can build more efficient and intelligent business automation.