Masters in computer science with specialisation in computer systems. Courses include Advanced networks, Advanced computer architecture, Reconfigurable computing, Advanced distributed algorithms and data structures, Explainable AI, Web security, Designing code analyses for large scale systems and Model based systems engineering.

Ventured into a new chapter to establish an electronic toys and gaming utilities business

* Worked on migration of Amazon Machine Learning Service (AML) in AWS from old tech stack to new React based frontend . Owned the entire design, architecture, development, operations, monitoring dashboards and e2e testing for the new console. Migration process included analysis of the older console and its dependant services and carefully retiring unused ones and coordinating the effort with several teams. The new service was fully written in React and Typescript, Java on the backend and Ruby for the infrastructure.* Worked on the development of a new service within AWS SageMaker called as SageMaker ModelCards. Created initial design, API design, API development, Infrastructure as Code, Operations and Dashboards. SageMaker ModelCards is a user facing service that lets customers create and catalogue model cards for their machine learning models. Backend of the new service is written in Go language and is deployed in a serverless architecture in AWS.* Worked on the design and implementation for fixing a long term security vulnerability within AWS SageMaker. It involved researching and analyzing the scale of impact and potential customers who already are facing the issue.* Worked on AWS SageMaker Studio and contributed to several UI features in React.js and bug fixes within the platform.

* Worked on a health project in Angular, Java, C#, Python and AWS. It has a complex data driven architecture design and data flow scaling to several gigabytes per second . Have designed efficient architectures in AWS to support the scale of incoming data. A data stream using AWS Kinesis was used to queue incoming health data and processed them in batch using Lambdas. Upon processing by the lambda, the data could reach a data warehouse (AWS redshift) for OLAP, object store (AWS S3), sent to another queue (AWS SQS) for further processing or to another data processing hub (AWS EMR). For faster access during OLTP, the data in the data warehouse was copied asynchronously to another AWS RDS PostgreSQL database.* Worked in the development of a cloud security tool that analyses issues in cloud configurations in Go, Angular, Python and AWS. The tool has a huge database of potential vulnerabilities and is updated asynchronously using a separate service. The tool used a serverless architecture and used Go as the backend language. Once a user submits a analysis job, the service schedules a job which starts a Docker container asynchronously in AWS Batch which in turns performs the analysis using the database of potential vulnerabilities. The tool has a Microsoft Active Directory style authorization mechanism (RBAC) built into it.* Worked on design and development of a video conferencing and broadcasting product using WebRTC and Websockets. The platform can reach up to 200 simultaneous participants in meetings. AWS Chime SDK was used for meeting management. In order to ramp up participants count to 200, multiple tests were conducted for several weeks and continuously monitoring the service in an automated way using Python and JMeter. For live broadcasting, a media pipeline using AWS Media services was developed and live video is hosted in CloudFront CDN which supports HTTP Live Streaming protocol.

A software development internship which was done during bachelor's degree. Learned about professional software development and the tools and techniques which large organizations use to develop, deploy and manage software. During this internship, the project was to create an employee management software in Java, JSP, Javascript and PostgreSQL database and ultimately deployed it to AWS cloud platform.