Dwarkanath Prabhu

Work Experience

Kindle Generative AI team

Co-Sell with Amazon (API migration)

• Led the transformation from an S3-based data exchange system to a real-time API-based system for submitting sales leads to AWS, improving efficiency for 70 customers
• Designed and implemented a secure architecture using IAM roles, STS, Lambda functions, and EventBridge, reducing data processing time from 30 minutes to 6
• Integrated the new API to automatically sync customer-defined solutions to the database, enabling customers to select solutions from a dropdown when attaching them to sales leads, eliminating the need for manual entry
• Built CI/CD pipeline to deploy the new AWS services using Jenkins and bash scripts
• Created CloudFormation infrastructure as code template for a seamless migration for all customers
• Mentored 3 junior developers by familiarizing them with the codebase, reviewing their code, and answering technical questions

Co-Sell with Google (new platform development)

• Led the development of Co-Sell with Google enabling customers to share sales leads with Google Partner Advantage Portal (PAP)
• Developed a subscription in GCP to receive leads from Google PAP pub/sub topic and share with customers
• Enhanced security by creating a Java service to store and rotate API secret keys in GCP Secret Manager
• Implemented OAuth 2.0 login in Angular allowing customers to use their preferred identity provider for authentication
• Onboarded 10 customers facilitating the sharing of over $1.5 billion worth of leads

Co-Sell with Microsoft (vendor to in-house migration)

• Created a Flask microservice to listen to updates in Microsoft Partner Center via a webhook.
• Designed a REST API to perform CRUD operations on leads in WorkSpan from aforementioned Flask microservice
• Migrated over 50,000 objects for existing customers to new microservice from PowerAutomate

Monitoring and Incidence Management

• Resolved 100+ customer support tickets by analyzing logs and performing root cause analysis
• Created a slack bot to post error messages to each product's respective channels for customer service teams to monitor
• Devised a script to publish daily and weekly error reports from trace logs in BigQuery
• Prepared a runbook for resolutions to most commonly occuring errors thereby reducing incidence response time

Internal Automation Scripts

• Sales report schedule email service in Java using Sendgrid SDK
• Product configuration installations for new customers
• Identify and delete duplicate objects from DataStore (GCP NoSQL database)
• Automated testing of internal APIs using cURL

Virtual Conference Platform

• Spearheaded the execution of two virtual conferences, each hosting 15,000 attendees, from concept to completion
• Developed a web application for users to register and upload presentation videos to an AWS S3 Bucket accessed by over 100,000 users
• Automated Zoom Meetings creation using Zoom API for live sessions with access limited to authorized participants

Sponsor Registration System

• Created customized sponsor registration systems in ASP.NET with a SQL Server database for 10+ conferences
• Designed an intuitive booth selection interface using CSS Grid
• Developed a secure payment system using Authorize.Net for credit card transactions

Paper Management System

• Refactored legacy PHP code into Python for efficient reviewer assignment
• Led the development of the API enabling paper authors to submit copyright to IEEE
• Implemented a robust full-text search functionality for papers through a PHP-based API with AJAX calls
• Automated creation of post-conference proceedings using a Flask app served via nginx on Debian

Skin Cancer Detection Using Deep Learning (Computer Vision)

• Analyzed a labeled dataset of 10,000+ skin images to classify into 7 types of diseases
• Achieved validation set multi-class precision of 87.5% - a 25% improvement over baseline by using ResNet50 (pretrained on ImageNet), Data Augmentation and Weighted Loss to fix class imbalance

Power Generation Capacity Planning – Conference Paper (Optimization)

• Modeled and solved a bi-objective optimization problem using linear programming to minimize water utilization and cost for power generation in Texas with CPLEX in C++ over the next 25 years
• Built an app using Shiny framework for R and GIS data to display optimal solutions on a map

Projects

• Home print server: Installed CUPS on a Raspberry Pi, enabling successful remote printing from a mobile device
• Payroll Processor: Full stack app to upload payroll report (hours worked, pay rate) and get amount due for employees with HTMX and bootstrap frontend, Flask backend, and SQLite database
• Student grading system: Full stack app to submit grades for students with a React frontend, Node backend and MongoDB
• JokeGPT: Twitter bot to use OpenAI API to write jokes based on a given prompt

Education