“Gauging customer satisfaction magnitude via social media platforms”


We have had the opportunity to work with one of the largest automotive industries in the country. They desired a real-time system to track user opinions, identify their clients and their problems across various social media channels.

Gauging customer satisfaction is a big challenge for car manufacturers as they can’t gauge the magnitude of sentiment from the data collected across various social media platforms such as Facebook, Twitter, and Instagram. This data didn’t provide the entire picture as it was too general and didn’t account for the sentiments communicated through platforms, Leading to an inability of car manufacturers to identify the different problems experienced by their customers, which delayed the development of an effective solution.

With the application of NLP, these issues can now be resolved in real-time as the system keeps a close eye on all customers’ communications and categorizes them into pertinent categories so that the manufacturer can act right away. 

Proposed Solution:

Amlgo Lab’s proposed solution has been a key towards effective and successful end-to-end implementation of the desired system:


Step 1:  The historical raw data is stored in AWS S3. Also, future incremental data will also be stored in AWS S3.

Step 2: The data in AWS S3 undergoes ETL transformation. A Glue job script is run to accomplish this task.

Before performing ETL on the data, considerable time is spent on exploring and analysis of data. These insights were key to building the NLP models required for categorization of data.

AWS Sagemaker is a very useful tool when it comes to experimentation, development and testing.

Step 3: The intermediate files generated in Step 2 are temporarily stored in S3.

Step 4: The intermediate files from Step 3 are appended to the Redshift database, and then deleted.

Step 5: The Redshift database is used for further downstream tasks, such as Power BI.


By allowing the client to concentrate on tasks that require human interaction, the automatic tracking has significantly increased the productivity while also lowered operating costs. Much sooner than before, the client was able to find and close gaps in their service and product development.

Our end-to-end methodology also assisted the client in identifying unforeseen areas that needed additional development to guarantee maximum customer satisfaction at all business levels. It was quite difficult to correctly grasp the text due to the noisy quality of the data and the linguistic cues. Finding a rich manner to portray the text was a major focus of R&D.  The essential components of the solution were extracted using an ensemble of many models.