At Meta, I am part of the Rights Management team. We build tools for Businesses and Creators to manage their copyrighted content across Facebook and Instragram.

• As the owner of the Creator Experience pillar, I create features and tools for creators to manage how their content is affected by Rights Holders and their digital copyrights

• As Tech Lead, I led 2 SWE, designed and developed the Preemptive Copyright Check for Meta Business Suite. This lets creators check run a copyright check on their videos BEFORE they Publish. Meta's top 200K creators use this feature every day.

• Built the Music Revenue-sharing experience for creators on Facebook. They can now use Licensed Music on their content, and share revenue with the artists

• Built the end-to-end next generation Creator Experience for finding out and resolving copyright claims on their content using Support Inbox on Facebook. 50 million creators use this feature every day.

• Responsible for building data pipelines for analysis using disparate intelligence and healthcare databases

• Making sense of the data by mining concepts to build a knowledge graph - helping investigators of counterfeit drugs by discovering associations between their cases

• Named Entity Recognition using an ensemble of state-of-the-art models in NLP, such as Spacy, NLTK, and Stanford NLP, to extract dates, locations, persons, organizations, etc.

• Building a web application to bring the analysis to life with full-text search and real-time analytics, using React & Node.js

• Designed and developed a machine learning based WiFi-LTE handover system at Samsung’s Mobile Network Operations Lab

• Extracted time-series features from Wi-Fi signal metrics and modeled classification and prediction of WiFi quality and signal strength

• Used Python with open source libraries such as tslearn, scikit-learn, pandas to model machine learning solutions

• Developed analytics and DevOps dashboards for keeping track of initiatives in the company

• Created custom Kibana plugins integrated with an Elasticsearch database

• Won second place in the JPMC Global Hackathon with a submission of a machine learning based payments fault predictor. The predictor and its services were deployed as a cloud application to eliminate manual intervention for correction of faulty transactions in the bank

• Machine learning models trained in Python on a large data-set of transactions

• Worked in an Agile Development environment, with daily scrum meetings and two-week sprints.

Skills:

• JavaScript, AngularJS and JQuery used for the dashboard

• Wrote REST APIs for data fetch, update, deletion and more for the server

• User interface creation using CSS, HTML and Bootstrap

• Integration with Spring for transaction management as well as SonarQube for code quality management

Course: Neural Networks and Fuzzy Logic

Expanded the course material by creating content on PCA, SVM, and LDA

Took MATLAB sessions on fuzzy control, perceptron classification, and function approximation using RBFNN. Evaluated and graded course projects and held problem solving sessions for 130 students