Nagaraj G Email & Phone Number

Bengaluru, Karnataka, India

Working for AWS OpenSearch Service

Bengaluru, Karnataka, India

Worked for Amazon Prime Video Org1. Worked on X-Ray feature for live sports. X-Ray displays stats, plays, teams and fan poll questions. Designed and lead the implemention of Presentation Service from scratch which transforms raw live event data to displayable data format in UI via template engine Freemarker, does view diffing and sends delta widget updates to client devices, time synchronizes live stream with X-Ray data. It reduced time to onboard X-Ray on new sport by 58% and new leagues/tournaments within same sport by 80%, onboarding new widget by 40-50%2. Designed and implemented a central configuration store service to streamline config management and avoid duplication and changes of configs across 5 services. It also aids in updating configs for any contingency scenarios during the live game, thereby providing better control to manage service3. Lead the design and implementation of the most requested feature Scoreboard widget for Soccer. Raw live event data is transformed into SDUI based model and dynamic updates to widget is sent in real time4. Implemented the question widgets for FanPolls feature. Implemented SDUI logic for rendering different types of questions widgets based on the nature of data. This improved customer engagement rate on X-Ray and FanPolls feature and made the live event streaming a 2 way interaction5. Worked on creating CICD pipeline for services, writing Integration and canary tests, load testing and scaling of services, creating metrics, alarms, monitoring and business dashboards6. Lead the team in code and design reviews, managing and maintaining good operational excellence

Bengaluru, Karnataka, India

1. Reduced Ads expenditure cost by integrating Google Ads performance data to enterprise data warehouse for performing analysis and taking data driven business decisions2. Decoupled tightly coupled integrations by designing and implementing microservices architecture. This resulted in easier enhancements, scaling, testing, deployment, memory and CPU utilization3. Implemented SSO for Akamai campaign builder and edge worker playground applications4. Lead the team in design and implementation of Integration projects

Bengaluru Area, India

1. Reduced transaction processing time from 40s to 18s by asynchronous processing for a critical and frequently executed business tranasction. The time consuming and not immediatly required tasks was removed and executed in separate transaction using queue processing2. Built a real time integration using push integration via Oracle Advanced Queueing. This resulted in customer satisfaction survey data reaching to support team managers to quickly review and respond in case of average or low satisfaction scores and to serve them better in future3. Implementing demand gen and lead management logics using database triggers and transactions4. Built a web application for employees to submit their ideas and POC to leadership teams to get support and convert their idea into business for Akamai5. Built several poller/push integrations capable of syncing 12 M records between systems6. Built monitoring, alerting and dashboards for integration applications using Splunk

Bengaluru Area, India

I worked for "Enterprise Search Engine" team. I was responsible for few of the tasks which went live namely1. Improved search query results by crawling, indexing & tuning relevancy of 0.2 M Akamai internal docs2. Writing API's for content curation tasks3. Created UI for Admin to add people with their roles and access4. UI development for displaying search engine results and for like/dislike, comment on each of the search results

Bengaluru Area, India

I worked for team responsible for processing of images from live camera. I was responsible for below tasks,1. Data preprocessing of images. This lead to Increase of 6% training accuracy of deep learning algorithm2. Automated testing of deep learning algorithm on paramters like memory, CPU and GPU utilization, throughput when feed with images at varying rates / min, latency in live video streaming, stability against system crashes. This resulted in eliminating manual testing and its efforts