Amazon Managed Streaming for Apache Kafka

RetryInstallAgents Workflow-: Implemented a solution to automate installation of AWS vulnerability management services such as Chronicle, and Inspector to monitor Kafka cluster instances in the service accounts.• Utilized AWS Systems manager through DataPlaneCommandExecutionWorkflow, effectively reducing Sev3 tickets by 90%.• Designed and documented a comprehensive workflow, ensuring idempotency, to handle the installation process, thus minimizing manual intervention and potential user errors.• Developed a proof of concept (POC) on the local development stack, showcasing the workflow's functionality, and integrated it with the Create Cluster workflow to seamlessly push failed instances into SQS. • Implemented robust monitoring mechanisms, including custom alarms and dashboards, leveraging AWS CloudWatch and Carnaval, providing real-time visibility into workflow status and ensuring timely response to installation failures.• Reduced the total turnaround time by 99% to install agents on the failed Kafka cluster instances.• Used Java with serverless technologies, including Step Function, AWS Lambda, SQS, and EventBridge Rule.• Onboarded this workflow to the CreateCluster workflow and deployed the changes worldwide through all 7 waves.

Teaching assistant for course CSE214 - Data Structures under Prof. Ahmad Esmaili

T-Notifier (Customer Announcements Application): Worked as a full stack developer and created T-Notifier that aided and enhanced the process of customer communication for database disruptions. Achieved the KPI of 99.95% availability of databases. Supported 6000 Hana/Sybase database customers.• Created the tool’s Low level and High-level architecture designs• Automated the announcement process from Service Provider Cockpit to Cloud Availability Center and enabledsupport to publish multiple customer announcements in a single click• Devised alerting mechanism to notify DBOPS team in case of unattended or delayed tickets• Included feature to generate reports for managers and stakeholders to make intelligent decisions• Enhanced the tool’s scalability and reliability through microservice architectureService Onboarding App: Worked with the TICO Cloud-Ops onboarding team to understand the pain points in the current onboarding process. Responsible for streamlining the entire process from service requirements to go-live. Enabled 2x faster onboarding with a strong emphasis on quality compliance.

Bill of Quantity Analytics (BoQ): A Web based application used to predict domain-specific 'Bill of Quantities' for a newly entered set of project attributes.- Data science: Machine learning models were trained on the historical data collected from engineering teams of the 'Power and Gas domain'. The training dataset consists of 60 different attributes and 100 target variables for the prediction.- Backend: MVC architecture was followed in the application development. REST APIs were created in Python-Flask and exposed to the front end.- Frontend: Angular was used to develop the User Interface along with other UI frameworks.Context Analyser: A Web-based application used to analyze and extract useful information from large site commissioning documents.- Machine learning models were trained with the data extracted from existing site commissioning documents.- Backend: Trained models were used to make a prediction for the newly entered document. REST APIs were created and exposed for data communication.Technology: Python-Flask, PyPdf, Machine Learning- Frontend: Angular framework was used to develop the User InterfaceSite Tool Tracker: A Web-based Application used to deliver & and keep track of 'Site Commissioning Tools' at Siemens Orlando office, USA. - Application aims to deliver tools on time without delay and also allows tracking of the items for effective operations.- Java Spring boot was used at the backend to create microservices along with Postgres as a database.- Angular was used to develop dashboards for different user roles and rich user experience.Krawal Visualizer- Created a pluggable web-based tool to visualize complex hierarchical data- Analyzed data patterns and changes for a better user experience- Developed UX tests to test the usability of mobile or web device- Utilized D3.js and Angular framework to generate dynamic graphs and charts- Developed RESTful API in Java Spring boot to access and use data

• Developed plant intelligence options: Performance Insight and Calendar• Performance Insight option enabled users to calculate key performance indicators and display using WinCC controls• Implemented various charts to demonstrate performance, quality, and availability of the production plant• Charts created for KPI indicators: Bar charts, Gantt charts, and Multiline diagrams• The calendar option allowed users to create and visualize structured planning of the production process• Used to define running times of machines and to create standard templates for regular production days and shifts• Utilized Java, Javascript, and Angular to develop the PI options