Globalization is everywhere.

Big companies have distributed branches, but how management can be sure that customer experience at high level in all locations? We have decided to check opinions about different branches of insurance companies in the USA.

What it does

Our POC suggests such flow:

  • Customer enters search string for locations and just wait for results 
  • Our background processing finds all related locations with reviews on Google maps. Then apply expert.ai NLP API Full and Classification analysis 
  • After that Status for current became "Finished" and the customer can check result on charts 

How we built it

For the demo, we've gathered reviews from all available locations for few companies. After processing reviews thru expert.ai NLP API, we've noticed a few things.

  1. Insurance agents have higher overall sentiment than, regular offices
  2. Geographical distribution can differ from coast to coast
  3. Some commercials made people angry, this can surprise marketing departments :)
  4. Entities from expert.ai allow to catch comments, addressed personally
  5. Classification from expert.ai allows finding cases, where customers addressing the area of problematic topic(Fire, Flood, Crime...)

Challenges we ran into

  • find interesting insights in a new data
  • connect data from differnt api
  • build data pipeline in short time

Accomplishments that we're proud of

  • we have build a new team :)
  • tried new technologies
  • found unexpected insights(complaints about commercial)
  • build POC

What we learned

  • expert.ai api shows relevant results
  • sometimes positive review contains negative sentiment(happy with service, but angry about commercials)

What's next for Places emotion

  • we'll build MVP

Built With

Share this project:

Updates