A company launched a new product two weeks ago and they hired a machine learning specialist to design a new system to detect the sentiment of their customers’ reviews automatically. The company does not have a labelled dataset for the specialist to train on. The company insists that the specialist should come up with a solution as soon as possible and that they have no intent to buy labelled data. The specialist is provided with a csv file on an S3 bucket in which all the reviews of the new product reside. The specialist should devise a solution that: Analyzes the sentiment of the reviews and dumps the output in the S3 bucket. Defines the schema of the data and runs SQL commands to clean the data if necessary, using a managed service. Creates a pie chart using a business intelligence tool to segment the reviews. Which architecture should be implemented?