Architecture of the wingbot.ai platform

Bubbles

We have come a long way in the development of chatbots. Different clients came with different needs. Thanks to our eagerness to fulfill their needs, we built a rich chatbot architecture that underpins what Wingbot is all about. Our solution can handle everything from integrations to various platforms, through a wide range of plugins, multilingual work with NLP to the orchestration of multiple bots at the same time.

Azure BotService
MS Teams
Slack
WeChat
Twilio / SMS
Alexa
Web chat widget
Fullscreen / Floating UI
(responsive customizable design)
Variable input
(password, text, buttons…)
Rich attachments
(with file upload)
Interactive web integration
(page-chat connection)
Voice
VoIP SIP integration
(telephony with handover support)
STT / TTS
(Azure/Google speech services)
Facebook, Instagram
Push notifications
(one time / recurring)
Facebook rich templates
(carousel, buttons, …)
Messaging orchestrator
User authentication
(SOAP, OpenID, OAUTH)
Omni-channel hub
(one chatbot connected to multiple channels)
Human Handover API
(with history management and authentication)
Multi-bot assistants
(master-bot architecture)
Live Chat Ready
Smartsupp
Genesys
Daktela
Natural Language Processing
Named Entity Recognition
(2-step fuzzy and strict entity matching)
System Entities
(@date, @num, @email, …)
Multi language intent recognition
(pre-trained language models: cs, de, en, sk, …)
Compound RegExp entities
(combines system and named entities)
Automated NLP training
(re-train with one click)
Training data analysis
(automated, static & cross validation)
Chatbot Application
Conversational Engine
free-form AI conversation
with context
decision tree dialogues
(3-layer fallback)
disambiguation
(for poor utterances)
FAQ Bounce
(dynamic digression)
thread state variables
(digression)
data anonymization
(implicit/explicit)
Rapid Plugins
mailing
SMS verification
input validation
custom integrations
(SOAP, REST)
Office 365
Notifications
(campaigns, tagged audiences)
Audit Log
(API Access, Conversation events)
Reporting
(Azure Table Storage, Google Analytics)
Chat Logs
(conversation history)
GraphQL API
(customizable, asymmetric JWT authorization)
Bot Training
Clustering
(topic discovery)
Maintenance reporting
(conversational recommendations)
Business Reporting
Data Studio
(Google)
PowerBI
(Microsoft)
Wingbot.ai Designer
Bot Manager UI
(notifications, deployments, history)
Dialogue Editor
real time collaboration
with tasks and comments
quick navigation
visualisation
of decision tree
drag & drop / copy-paste
multi environment
versioning
Rapid Plugins
SSO login
2F authorization
role based access
audit logs
Tester UI
(automated conversation testing)
NLP Trainer UI
(intents, entities, validation)