130 likes | 146 Views
Michal Pomykala VIRTUAL ASSISTANTS OUR JOURNEY FROM EXTENSIVE DOCUMENTATION AND VARIOUS TOOLS TO SMART CHATBOTS code::dive 2018 Lunch talks 13:35-13:50, stage 9. our journey. research. our journey. documentation. our journey. trainings. our journey. analytics. our journey.
E N D
Michal Pomykala VIRTUAL ASSISTANTS OUR JOURNEY FROM EXTENSIVE DOCUMENTATION AND VARIOUS TOOLS TO SMART CHATBOTS code::dive 2018 Lunch talks 13:35-13:50, stage 9.
our journey research
our journey documentation
our journey trainings
our journey analytics
our journey dedicated tools
our journey but still.. is it enough? applications share point specification wiki teams trainings analysis …
our journey so let’s try with chatbots and AI/ML
design all in one… if possible … documentation trainings analytics tools research SMART CHATBOTS
chatbot basic architecture services integration additional chatbot logic NLP/NLU - human understanding user platform databases analytics knowledge dedicated tool and API internal services
what is the user intention? domain knowledge need support dialog I would like to know … repair a bike intents order a pizza recognize what your users want set up meeting fallback intents .csv follow-up intents output context Q & A
what are your dialog components? system entities* @sys.any @sys.number developer entities** required? @sys.time user entities*** entities @sys.date conversation parameters @sys.currency @sys.places @custom.entity2 @custom.entity task=“deliver_pizza" pizza_name=“$custom.pizza_name“ pizza_size=“$custom.pizza_size" location=“$sys.address" *** login, session, orders ** custom, dedicated for service * available and define in platform