Product:
FMCG Chatbot and Neural Net
Agency:
Jubi.ai
Role:
Business & Product Manager
Problem
FMGC, companies globally serve a very segmented customer base and it is especially true when it
comes to enterprises, and due to the nature of their business they face a unique issue specific
to their sector, where in the range of the queries and grievances is vast and customer segments
are often very different as products offered in the case of enterprises is huge.
Challanges
- NLP model & Neural Net needed to be trained on FMCG.
- Due the nature of business the users bases were a mixture of urban and rural crowed
- The QAs had to take care of a number of sectors 6 to 7 everyone drastically different from
other
- Each line of business had different requirements and needed different ways of delivery for
their respective customer data.
Solution
- India’s First FMCG trained neural net with the ability to add client specific questions.
- 6 unique flows to cater to 6 unique lines of business.
- A delivery mechanism to send the data from each line of business to the relevant teams, via
secured emails.
- In a First, multi-personality conversational interface, to make the experience more
relatable for different customer segments.
- A PWA that could run on internal data centers, with a single api connection to the NLP
engine.
Result
Automated about 10% of their yearly queries via the interface.
