My background/strengths are in analyzing bottlenecks and inefficiency up and down the stack, and principled, data driven engineering to improve the underlying systems. In the recent past, I have worked on:
1. Improving video streaming on Meta's apps
2. Accelerating web applications and Ad stacks, at scale
3. Large scale, low-cost video delivery on wireless data networks
4. Inferring performance bottlenecks in Tier-1 IP and 4G data networks.
I also have considerable experience in processing large amounts of data through its life-cycle - from ingestion/storage at fiber-link speeds, building aggregation/ETL pipelines and developing inference/analysis techniques.
1. Improve video streaming performance on Meta's Family of Applications
2. Optimize Meta's Edge Network Deployments for Application performance.
Researcher in a Large scale, low cost content delivery networking group. Projects include:
Detecting TCP/IP anomalies at line-speed in the Sprint Tier-1 network backbone.
Realistic work-load generation for web-server load tests
Web, ad-stack and network performance analysis
Data analytics and big-data platforms (HDFS, Hive, Spark, ElasticSearch)
Network protocols and architectures (both IP and Wireless data networks)
Modern web and browser technologies (HTML5, Web workers, Web streaming protocols, Http2 etc)
C, Typescript, Java, Python, Unix shell, SQL, R
Nginx, NodeJS, Linux, Kubernetes environments
Monitoring and alerting systems (eg Prometheus, Telegraf , Nagios, Graphite)
Publish-subscribe and messaging systems (Kafka, ZeroMQ, RabbitMQ)