Dot Bot Use data for 100 Problem Call
Dot. Bot Use data for 100%
Problem ● Call center congestion ● Companies accumulate a lot of data every day but do not work with them ● Shortage of personalized content Data is ore. If you have better tools, then you mine more gold
Dot. Bot Tool based on AI Unload call-center + 3 stages answers which prevents mistakes. Near future plans + Voice bot will replace 1000 operators + Robotization of call centers Facebook page data gathering (graph API) +Likes +events +significant dates +friends +Photo tags +Posts Portrait of client • • Personalized content / offer Prescoring Cross sells Clients gets answer immediately
Monetization B 2 B, particularly for banks and retailers $ $ $ 1 2 Fees per month depend on quantity of sessions Fees per successful session • Banks: prescoring or bot sold credit, deposit • Retail: Cross-sells based on user events
Companies that interested in our product
Thank you for your attention Arkhat CEO & developer Bekzhan Developer Darkhan Developer Yersain ML engineer at Driven data advantages
3 stage answer First of all chat bot train for instance 100 question types and 1000 variations of each question (By training this number will be continuously growing) If Dot. Bot pretty sure what question means then answers automatically If Dot. Bot not sure for 100% then give his options what the question means to passive operator After every answer Dot. Bot become better and better If question not exists in variation of questions then active operator will chat with client
Analysis of solution After implementation of VS our product Now • Client waits long time • Save on call duration • Permanent shortage of call center operators • Dependence on call center staff • No connection between different source of communication • Accumulate a lot of data but do not use them • Immediately answer from bot • There is no time limit, the bot can correspond as much as you like. • Dot. Bot replace operators • Dot. Bot works 24/7 and always in the good mood • All chats in one place • Dot. Bot remembers client questions and after analyzing a profile prepares personalized content.
Implementation process 1) Get from Company list of FAQ and scripts of their call-center (1 week) 2) Train bot based on this data (1 week) 3) Give them admin panel where all chats will in one place 4) If company does not have current chats (FB, widget, telegram etc. ) we create for them (1 week)
Training of bot Dot. Bot answer to client automatically Dot. Bot send the question to silent operator and propose answer Live operator Human chats with human and bot Silent operator correct suggested answer and by train from results of conversation confirmation train the bot
Example of answer
Analyzing Facebook pages Likes Posts Friends Significant dates FACEBOOK Graph api Photo tags Events
- Slides: 12