An analytical platform for estimation and analysis of cost expenditures on the Amazon Web Services platform. Included:
1)resources monitor, which queried all available user resources and their utilization (server, databases load, etc), based on a predefined schedule
2) cost estimator, estimated daily, weekly and monthly projection based on current cost tariffs
3) recommendation system which helped to cut down expenditures. Architecture based on LSTM neural network and additional custom logic.
Python backend, with the utilization of AWS infrastructure (Dynamodb, S3, Lambda and Step functions, SQS, SNS).
Python 2.7, PynamoDb entity framework, Log4j, python multithreading and multitasking libraries, AWS (DynamoDb database, Lambda, Step functions, SQS, SNS, Pipeline, MQ), AUTH0