Intent/problem recognition (via Natural Language Processing NLP):

• Recognizes meaning of consumers’ questions, typed in a natural language, in a sentence, without the need for typing keywords, understanding synonyms of queries, correction of typos, language errors.
• Contextual Autocomplete: displaying intelligent suggestions whenever user enters a query in the search box, suggesting the closest hit response
• Automatic suggestions: are the result of analysis of previous questions – Manual suggestions: are the result of setting the desired suggestions (marketing information, targeting).
• Voice questions: integration with speech recognition system – ability to include functionality recognized voice commands (and not just typed queries) via API

Providing instant answers:

• Generates instant answers to users’ questions and /or running
  a virtual conversation from the Knowledge Base of Automated Answers, consisting of questions and answers dedicated to a given implementation, built:
• in a manual manner – by the team of Knowledge Engineering, based on customer materials
• automatically – the ability to index the speci ed web page (own crawler and indexer) and automatically build a question and answer database,
• based on data generated by external data systems
• Display Suggested Replies: Most relevant in a given context 3 – 5 replies or related topics

Answers in the form:

• Displaying text responses
• Display graphics (eg instructions, pictures, photos) in the content of speech
• Display multimedia, including instructional videos, in the content of speech
• Display site links in the content of speech
• Display search results from other search engine – web (eg Google, Bing)
• Emoticons display (for deployments in selected social portals)
• Display search results from other search engines on the company site / portal (from CMS systems).

Multilingualism – recognition a chosen language through

• use of automatic translators or dedicated semantics (eg English, Polish…)