Build document-based knowledge
Register PDF, TXT and image files to build a knowledge base from which AI can draft intents and answers.
Cloudturing features
Connect documents, tables and APIs with
Connect documents, tables and APIs with
AI chatbot creation and operation tools
Cloudturing goes beyond uploading business materials. Manage document-based intent generation, business integrations based on API documentation, table data search, security and operational logs in one console, from chatbot creation to daily operation.
Integrate using API documentation
Use your API documentation and schemas to prepare chatbot integrations for lookups, requests and authentication.
Search table data
Register product, branch and policy data maintained in tables and use it with Meilisearch-based search.
Operations, security and analytics
Manage business chatbot operations with member permissions, IP allowlists, usage, conversation logs and audit logs.
Workflow
From uploading materials to improving operations
The workflow is designed for operator review, combining the convenience of generative AI answers with reliable business support.
Upload materials
Register documents, images, API documentation and table data as chatbot knowledge.
Generate AI drafts
AI generates intents, draft answers and categories for operators to review.
Configure operations
Adjust the persona, design, security, member permissions and plugins.
Deploy and improve
Run the chatbot through a web widget and KakaoTalk, then improve it using logs and dashboards.
Knowledge builder
Core tools for document-based AI chatbots
Register materials for business support and let operators review AI drafts to manage the quality of actual answers.
Manage document files
Upload PDF, TXT, JPEG and PNG files and manage knowledge materials while checking file counts, storage and token usage.
Set the chatbot persona
Set the chatbot's name, role, tone and knowledge scope, and refine persona drafts with AI.
Generate conversation intents
Generate intents and draft answers from selected documents, and adjust category creation and generation options.
Edit and review answers
Operators can directly edit AI-generated intents and answers and manage them for actual chatbot conversations.
Data & integration
Connect answers to business data
Build business chatbots that use external systems and table data beyond simple FAQ answers.
API integration
Business integration from API documentation
Register API documentation files to help AI prepare function descriptions and call structures for integrations. Use your existing APIs for service requests, order or booking lookups and user authentication within chatbot workflows.
- Upload and manage API documentation
- Support for organizing function descriptions and call structures
- Configure internal user authentication and external API-based authentication
Table search
Precise table data search
Use data maintained in rows and columns, such as product specifications, price lists, branch directories and policy tables, as chatbot knowledge. Meilisearch-based search can find relevant items to include in answers.
- Use table data such as CSV and XLSX files
- Precise search based on rows and columns
- Suitable for product, branch, policy and internal information
Operation console
Manage business operations in one console
After building a chatbot, manage the settings, permissions, logs and statistics that operators review each day.
Design and channel settings
Manage widget colors, icons, header images, KakaoTalk chatbot visibility and QR code settings.
Install a web widget
Installation JavaScript and a web URL help you deploy a chatbot on your website.
Manage members and permissions
Invite team members and distinguish owner and member permissions to manage chatbot operations as a team.
IP access control
Restrict access to the chatbot management interface using allowed IP rules to strengthen internal operational security.
Usage and conversation logs
Review usage by channel, key intents and full conversation histories, and download CSV files for analysis.
Audit logs
Review changes to security, plugins and settings in audit logs to improve operational traceability.
Plugins
Enable and manage the plugins your workflow needs
Manage user data storage, user authentication, API calls, internal user authentication, error notifications, satisfaction surveys and watermarks in each chatbot's settings.
- User authentication
- API calls
- Internal user authentication
- Error notifications
- Satisfaction surveys
- Watermark
Feature FAQ
Questions to consider before adopting the features
Review the feature scope and operating approach when connecting a document-based chatbot to real business workflows.
What materials can I register as chatbot knowledge?
You can register documents and images such as PDF, TXT, JPEG and PNG files, and manage tabular materials separately as table data. For API integrations, register API documentation or schemas as separate materials to help prepare the integration.
Are AI-generated intents and answers published immediately?
AI generates intents and draft answers from uploaded materials, but operators can review and edit them before using them in actual chatbot conversations. This workflow helps reduce wording that does not match company documents and manage answer quality.
Does uploading API documentation complete every integration without development?
API documentation helps organize function descriptions and call structures, but not every environment is completed automatically. The customer's API authentication, firewall, test data and call permissions must be checked against the actual operating environment.
Can the chatbot handle service requests or internal authentication?
If the customer has request, lookup or authentication APIs, they can be connected to chatbot workflows. This suits scenarios that require business actions after an answer, such as service requests, order or booking lookups, stock checks and internal user authentication.
How does table data search differ from ordinary document search?
Table data tools handle materials where rows and columns carry meaning, such as product specifications, price lists, branch directories and policy tables. Cloudturing uses Meilisearch-based search to find relevant table entries and include them in chatbot answers.
Where can I deploy the chatbot?
Installation JavaScript and a web URL are available for websites, along with settings for KakaoTalk chatbot visibility and QR codes. Operators adjust design and channel settings in the console to suit their brand and usage environment.
Are team operations and security settings supported?
Member invitations, separate owner and member permissions, IP allowlists and administrator audit logs support team operations and access control. Usage and conversation logs help track operational quality.
What features do plugins provide?
Manage user authentication, API calls, internal user authentication, error notifications, satisfaction surveys and watermarks for each chatbot. Operators can enable the business features they need and check their status in the console.
Design a chatbot around your existing materials and APIs
If you already have documents, tables and APIs, you can organize them into chatbot knowledge and business integration workflows in the Cloudturing console.