Milestone 1-Build a model inference service

[danyatingshen]

Master issue:

Task summary: https://github.com/danyatingshen/Group3-Movie-Recommandation-System/issues/1

Build a model inference service that provides movie recommendations on request given your learned model (e.g. using flask). Specifically, we will send http calls to http://:8082/recommend/ to which your service should respond with an ordered comma-separated list of up to 20 movie IDs in a single line, from highest to lowest recommendation. We will wait for answers to our requests for at most 800ms. You can recognize whether your response has been correctly received and parsed by a corresponding log entry in the Kafka stream (expect to see an entry ,, recommendation request , status <200 for success>, result: , ).

Step:

• [x] Create a working flask basic service: https://github.com/danyatingshen/Group3-Movie-Recommandation-System/pull/4
• [x] Deploy to VM: see doc under VM operation and Kafka session
• [x] Serve the service on VM and make it externally available: https://github.com/danyatingshen/Group3-Movie-Recommandation-System/pull/5
• [x] Add get recommendation API call: https://github.com/danyatingshen/Group3-Movie-Recommandation-System/pull/7
• [x] Read from Kafka
• [ ] Write back to Kafa

[sakshik-cmu]

Write back to Kafka: Not required