Python + AWS + Statistics + Machine Learning
Budget: ₹1,500 – ₹12,500 INR
Problem Statement: My business partner is using AWS EC2 spot instances to run several jobs but many of these jobs are not utilizing max usage of CPU or memory of the instances. So there is a possibility of saving the cost by suggesting an ideal EC2 instance type before running a job.
So I want to create a recommendation engine that can recommend an ideal AWS EC2 spot or on-demand instance type for a job.
Currently, I don't have the exact customer's environment data (AWS EC2+Cloudwatch+Jobs data) but I do have dummy records for the same. I also have a few rules to validate the result of the recommendation engine. So the final solution may be the combination of the rules + Machin Learning.
In the end, the recommendation engine will be deployed on the customer's environment, and for a particular job it should recommend an ideal instance type.
So I want to create a recommendation engine that can recommend an ideal AWS EC2 spot or on-demand instance type for a job.
Currently, I don't have the exact customer's environment data (AWS EC2+Cloudwatch+Jobs data) but I do have dummy records for the same. I also have a few rules to validate the result of the recommendation engine. So the final solution may be the combination of the rules + Machin Learning.
In the end, the recommendation engine will be deployed on the customer's environment, and for a particular job it should recommend an ideal instance type.