johnsonlarryl
commented
1 year ago Hey @Monishgalla. Here are some follow up work that I would recommend:
- Do more data science/validation using various lags (ie. 30, 90, etc...).
- Do more data science/validation on what historical data should be used instead of 2012-present.
- Do more data science/validation by including more ticket symbols.
- Deploying the model to AWS using CloudFoundry (CF) as Infrastructure as Code (IaC) and Elastic Container Service (ECS)
- Adding more unit/integration testing within the application code
- Perhaps see how much faster compiling the code using Cython to speed the training and/or deployment.
Wrap up with achievements once again (very brief), talk very briefly of follow up work. Do not exceed 3 minutes.
@johnsonlarryl and @moalim02 put comments in this issue to give @Monishgalla some additional talking points for add on work.