The following repositories need to be downloaded:

• Whisper Api: https://github.com/CoveoWhisper/WhisperAPI
• NLP Api: https://github.com/CoveoWhisper/NLPAPI
• ML Api: https://github.com/CoveoWhisper/MLAPI

1. Create a Dialogflow agent: https://dialogflow.com/
2. Generate dialogflow_secret.json (see also: https://dialogflow.com/docs/reference/v2-auth-setup) To generate the file, go on the Google Cloud Platform/Service Accounts page for the project: https://console.cloud.google.com/iam-admin/serviceaccounts. Then Create Key and choose the json format. Then rename the file as dialogflow_secret.json and copy to the /api directory
3. Generate the query_model.bin file and paste in the /api directory (see Generating Models)
4. Set port in index.py

5. In a command line prompt run the following command: pip install -r requirements.txt

1. Generate facets.bin and copy to root folder (see Generating Models)
2. Generate documents_popularity.json and copy to root folder (see Generating Models)
3. Generate documents_searches__mapping.json and copy to root folder (see Generating Models)
4. Set port in app.py

1. Create appsettings_secret.json file at the same level as appsettings.json with the following content:

{
  "ApiKey": "yy50124000-aaaa-8888-dddd-0003dc9ec111",
  "NlpApiBaseAddress": "http://localhost:5000/", 
  "MlApiBaseAddress": "http://localhost:5001/",
  "SearchBaseAddress": "https://platform.cloud.coveo.com/",
  "OrganizationId": "orgId2940293"
}

• ApiKey: should be set to the Coveo Search Api key.
• NlpApiBaseAddress and MlApiBaseAddress: should be set to api addresses from 2.4) and 3.4)
• OrganizationId: should be set to the Coveo organization ID.

2. The app runs on http://localhost:52256/

http://localhost:52256/swagger/index.html/