Prasanth Ravulapalli

A microservice-based architecture. I designed the entire architecture and built the back end on Django and a few lambdas on JavaScript and Python. A Neo4j clustered database is used as a scalable graph database where queries are optimized to last from 40-50 seconds to 0.1 milliseconds. I used AWS Fargate to deploy the back-end microservices and AWS lambdas as cost-effective, scalable solutions for long-running tasks. Finally, I implemented AWS code pipelines to have proper CI/CD and to deploy each environment accordingly, as well as reliable and efficient ETL pipelines that accept 2-3 million data recorded daily. The whole infrastructure is managed using Terraform scripts for better deployments.

A web app developed using Django and the Flask framework. All databases, including psql, Neo4j, and Google Bigtable, scrapers, content processors, and queuing infrastructure are maintained on Google Cloud servers. Tasks related to the platform's intelligence were developed using datasets created by crawling over the internet, which grows at the rate of 50-100 thousand new entries per day, building respective data pipelines. I maintained a Neo4J schema to efficiently create clusters from the data, as well as multiple Bigtable instances to find proper indexing or partitioning of data for efficient search results. Finally, I developed and published a Chrome extension to create more data points.

Created a process that involves frequent scraping of Weibo and Twitter to gather data on selected users, their connections up to 20 levels deep, and their platform interactions. This information is utilized to create network clusters and gain insights into community behavior. I enhanced the system by developing dashboards using ChatGPT APIs, enabling dynamic query generation and data analysis. I considered optimizing scalability and efficiency, ensuring data quality and privacy compliance. I also leveraged advanced Neo4j DS analysis techniques for deeper insights and implemented real-time data collection, interactive visualizations, and a feedback loop for continuous improvement. I worked on regular maintenance and adaptations to evolve the product.