Model can be downloaded from:-https://drive.google.com/drive/folders/1neHfilNkrCwtry3dA7bTCADH3EheKdYo?usp=sharing
This is a question answering system which is used to service users at websites.
- It is quite often that we hire some people as customer service so they can interact with user's when they come on the website but what if we can automate this task?
- Because users usually come can ask similar questions and have similar requests a majority of this work can be automated.
- We create Megatron a question answering system.
- It can:
-
- take input from user in text, audio or video format
-
- classify user requests and provide the respective service or information.
-
- If the user is not satified then can let someone attend the user.
- To automate customer service.
- Python programming.
- NLP
- Data Validation
- Data preprocessing
- feature engineering
- model creation
- logging
- model fine tuning
- model selection
- Industry standard OOP's based approach
- logging.
- clustering methodology for better model building.
- data science lifecycle
- The various nlp architectures like bert, GPT, albert, distilbert
- deployment on aws.