Deep learning dialogue engine with outstanding performance

· Deep learning: deep learning model based conversational engine, requiring no feature engineering resulting in more intelligent chat than rule-based dialogue engines.

· Dialog control: Powerful dialog control module supports flexible intent switching.

· Industry leader: leading in recognition accuracy on public test data sets across multiple industries such as insurance, pharma, retail, telco and more.

Graphic building experience without coding, making business people masters of robot creation

· Visualization: Build multiple rounds of dialogue using graphics allows robot-building by creating flow charts.

· No code: Automatic knowledge point generalization tools enable business users to maintain Q & A knowledge bases without scripting or coding.

· Industry customization: Custom libraries maintained for industry specific nouns and synonyms.

· Instant review: Robot debugging window allows business people to conveniently test the effect of dialogues.

Comprehensive dialogue control capabilities make robot replies more unified and more compliant while creating better memories

· Natural Language Understanding (NLU) supports converting natural language into search conditions and retrieving structured or unstructured database content such as document knowledge base, business system database, knowledge map, etc.

· Document understanding: Supports direct document structuring for document-based Q & A without editing the knowledge base.

· Knowledge Graph: Make those sentences that have the same intention but are associated with different entities no longer confusing in terms of intention with the help of two-dimensional tables, and find the corresponding answer by reading the table.

· Automatic clarification: When a user's intention changes, the robot will complete the intention switch by providing the user with clarifications.

Powerful dialogue effect evaluation and BI capabilities make online dialogue effect optimization extremely easy

· Data monitoring: Supports continuous data monitoring and customizable report presentation, meets the data analytics requirements of different industries.

· Knowledge discovery: Supports automatic clustering of semantically similar problems to automatically learn and discover new knowledge points.

· Self-test and self-evaluation: Support automated knowledge base testing and health checks, help operation personnel find and fix potential problems of knowledge base in advance.

· Continuous learning: Supports online annotation and continuous learning, improving robots' performance via human computer interactions.

Support multi-language, rich media interaction, and personalized responses, and make dialogues colorful and individualized

· Multi-language interaction: supports 90 + dialogue languages and fast language expansion.

· Rich media interaction: supports rich media interaction methods such as text, pictures, videos, files, hyperlinks, etc.

· User profile: supports providing users with recommended replies with different content according to different user profiles, thus enabling an exclusively personalized question-and-answer experience.

· Personalized experience: Obtain and apply user attributes in various ways, allowing flexibly configured dialogue scenes such as personalized welcomes, recommendations, searches and pushes, etc.

Provide rich, open APIs, platform access and upgraded expansion capabilities

· Multi-channel: supports flexible access methods.

· SDK provides access via Whatsapp for Business, Microsoft Teams, FB Workplace, Twitter, Line, Twillo, Web, Avaya, Genesys, etc. Users can click or scan the HTML5 trial page to quickly experience the robot capabilities.

· Easy integration: Supports flexible extended development.

· Docking with other mainstream enterprise management systems supported --- ERP, CRM, CMS, etc. (Sharepoint, Slack, ServiceNow, SalesForce, etc.)

Enterprise-level security to keep robots continuously online

· Flexible permission configuration system fully meets the requirements of enterprise hierarchical permission control.

· Supports public cloud, private cloud and local deployment of multiple deployment methods.

· Cloud service MTBF > 99.99%, load balancing mechanism of cluster private deployment virtually eliminates downtime.

· System passes third-party permeability safety test.