• Built a robust and scalable in-house Payments Processing & Subscription Management system "Charon" for 3 tiers of service at Pandora - Ad Supported, Plus and Premium

• Charon helps manage subscriptions and payments from 5M Pandora subscribers with revenue > 300M.

• Charon deals with Apple, Google, Amazon and Roku subscriptions, along with Paypal and Chase Paymentech credit card transactions

• Helped build batch threads to renew Apple/Google subscriptions and revalidate in-app subscriptions by verifying the Google and Apple receipts.

• Built batch tool "GoPA" to publicize Premium by offering millions of listeners complimentary Premium for a week without any sign up or credit card information, yielding 700K trial starts.

• Built scalable backend APIs for a viral link that would give listeners free Premium for 3 days - www.pandora.com/trypremium

• Ensured CCPA, PCI and SOX compliance on all our payment servers.

• Built a customer service tool with a whole suite of functions to resolve listener's issues received through Salesforce Helpdesk.

• Developed tools for marketing and customer service teams to deal with redemption and generation of Gift codes & Promotional codes

• Dealt with deployment of VMs, Fraud check and customer service support as part of my on-call responsibilities

• Helped Pandora with our 2017 rebranding initiative with new logo and graphics

• Exposed to new technologies such as - Kafka, Redis, Postgres, Handlebars, Apache Airflow

• Developed Pandora's email delivery microservice "Hermes" to send millions of emails through Sendgrid

• Built an in-memory cache of licensing rights of songs to determine whether a song could be replayed, skipped, played on-demand or downloaded

• Worked on Airflow DAGs using Sqoop to import listener state snapshots into Hive and export apple transaction report by querying Hive on Hadoop cluster

• Participated in hackathons -1800-Pandora (Twilio API based phone dial service to listen to Pandora radio) and Trendora

Fraud Management System PL/SQL, Java, J2EE, JSP, JavaScript, Struts, JQuery, AJAX

• Developed a backend system to identify potentially fraudulent bookings on Priceline.com, which are then worked on by call-center agents across the globe to confirm whether the booking was indeed fraudulent.

• Developed a system of Java batch jobs and improved an existing Java Struts app to create new cases for agents to work on and close irrelevant cases based on fraud detection strategies popular at the time.

• Developed a framework consisting of PL/SQL stored procedures, Java application and complex tuned SQL queries in Oracle database to design and dynamically add/remove/update fraud detection strategies

• Developed front-end in JSP/Javascript for agents to get cases, view booking details and make calls to the bank

• Developed a J2EE based priortization engine which maintains a distributed in-memory cache across several servers to dispense cases to agents based on their skillset, potential dollar savings on the booking and call center location

Integration of Accertify with Fraud Console Spring REST web services

• Developed REST web services to add/remove fraudulent transactions onto their consolidated database and ban/unban fraudulent users from booking on our websites in the future.

• Developed Java APIs to exchange booking information and Java batch jobs to sync our system with theirs.

Customer Care System Java, J2EE, AngularJS, JQuery, JavaScript, JSP, AJAX

Full-stack Customer Care projects – Spring REST web services & AngularJS/JavaScript/JQuery/AJAX front-end

• Built a J2EE application to process full/partial cancellation of insurance premium(s) on all bookings

• Implemented a system to give hotel a new credit card by utilizing REST APIs provided by third party corporate payment companies - TSYS/WEX to charge Priceline

• Built an interface for call-center agents to view current, pending and settled transactions on customer’s credit card

BI Dashboards Reporting Server - Java, JSP, JavaScript, AJAX, XML, BIRT

• Developed the Reports server which would periodically generate reports on charts or dashboards using Business Intelligence & Reporting Tools (BIRT), and deliver the reports to clients via Email or receive an alert over SMS

Flash Plugins for iDashboards enterprise software - Flash, Flex, ActionScript, JSON

• Developed flash plugins such as Analog clock (timezone supported), Digital countdown clock & Calculator and APIs to integrate standalone flash apps onto a dashboard

• Added proxy server into the cloud/SaaS edition for fetching images from specified URL and loading it onto dashboard.

• Developed color palettes for recently used colors (http://gallery.idashboards.com/preview/?guestuser=webhc&dashID=228)

Intra-company Management Systems - Java, JSP, JavaScript, AJAX

• Developed License Management System which reduced the potential client’s waiting time from 3 days to instantaneous.

• Streamlined the process of scheduling engagements & client surveys by developing an Engagement Management System

• Developed an Insurance Management System for the construction projects at MSU in Java/JSP on GlassFish server.

• Integrated this insurance system with Oracle Enterprise Project Portfolio Management using SOAP services.

• Developed PERL & Shell scripts to generate XML inputs for web services of Oracle Primavera over SOAP to automate synchronization of FAMIS accounting system payment records with Primavera records. (https://unify.gis.msu.edu/cms)

• Developed web content to describe the business processes of construction management using HTML/CSS on Drupal.

I was working as a Research Assistant in Electrical & Computer Engineering at Michigan State University. I was working on the project "Team cohesion and Conflict Management of On-board crew members." This project is funded by NASA and is the primary source of funding for my research assistantship.

In parallel with this, I was also working towards my doctoral thesis in the area of "Delay Tolerant Networks." My thesis aimed toward quick content dissemination in networks where the connections occur intermittently (e.g. a network of users carrying mobile phone over Bluetooth). The applications of this project would be of interest to content distributor companies such as Amazon.

Doctoral Research ­ (Projects funded by: NASA, NIH)

Human Interaction Detection

· Developed the code in TinyOS for Heart-rate measurement on a Cricket sensor node

Human Posture Recognition

· Implemented recognition of human posture on Mica2 motes in TinyOS

Energy expenditure for computation on sensor nodes

· Study of power consumption in sensor nodes through experiments and extensive literature survey