The AWS CDK Project was built via codepipeline for continuous delivery. A step function defined several steps including lambda functions that made API Calls, basis which further processing would be started (Including saving new driver details to dynamoDB table).

Services used: AWS Lambda, DynamoDB, SNS, SQS, Step functions, CodeBuild, CodePipeline, S3

(01/2021 - Present)

Created Tableau Dashboard to demonstrate the cost/usage of SageMaker service.

It also illustrates the total number of users and SageMaker notebook on a monthly basis.

This helps users determine how much usage and cost they are adding to their cost center on a daily/weekly/monthly basis

Tools and Technologies: Tableau, SQL Server.

(10/2020 - Present)

Creation of docker image for ML models that needed to be moved to production as well as creating lambda functions to run them.

Identifying base docker image as well as library dependencies.

Tools and Technologies: Docker, Python, AWS Lambda Functions, Go Lang, AWS SNS.

(10/2020 - Present)

To generate Cloud Native Utility to convert on- prem created Python scripts in accoradance to Cloud Environment as per SLF standards

Tools and Technologies: Python, AWS Sagemaker, EC2, Unix, PyCharm, S3, Jira

SageMaker Endpoint : Testing the SageMaker endpoint service and calling model from an outside AWS service (e.g. Postman).

Model Predictions to FTP Server : Built and tested binary file that saves model's prediction to an FTP Server that are ingested into Salesforce.

Tools and Technologies: Python, AWS S3, AWS SageMaker, golang, Cloudwatch logs

SageMaker Endpoint : Testing the SageMaker endpoint service and calling model from a service outside AWS (i.e. Postman).Model Predictions to FTP Server : Built and tested binary le that saves models prediction to an FTP Server that are ingested into Salesforce.