Bing Ads Campaign Plat Dev• Focused on performance optimization for a distributed system that loads millions of entities per request. (C#, SQL, Kusto)

» CiteSeerX | Feb 2020 - 2021• Co authored a paper published on IEEE: Building an Accessible, Usable, Scalable, and Sustainable Service for Scholarly Big Data• Worked full stack on the new generation of CiteSeerX, an Academic Search Engine, using Vue.js and Python.• Worked on document similarity prototype using doc2vec, BERT, co-citation, and bibliography coupling. Tf-idf from Elasticsearch’s similar document API is used as the baseline model.» Intelligent Vehicles and Systems Group | 2018 - 2019 • Researched EEG and EMG data’s correlation with decision spaces – branching areas of predefined paths – to assist with robotic wheelchair steering for patients with motor control deficiency. • Implemented classification pipeline of sEMG signals on forearm muscle to joystick control. A 2nd order bandpass filter with frequency bounds of 5 Hz and 450 Hz was used on the preprocessing stage. Found rolling variance was the most effective in feature extraction. Lastly,linear discriminant analysis (LDA) was used for classification on extracted features.• Implemented classification of EEG signals between rest, right hand, and left-hand movements with linear discriminant analysis (LDA) and support vector machine (SVM) using an online EEG motor imagery data set.(Python, Scikit-learn)

•Worked on Distributed Tracing with Open Telemetry to collect network traces and service metrics as an SRE intern.• Started the implementation of a new Open Telemetry feature for use within Capital One to detect gateways and third-party APIs. (Java)• Instrumented existing Capital One services to export traces to Jaeger and New Relic One.

• Developed a public API using .NET Core that optimized the customer’s production costs based on relational data. (C#)• Built many features for the public API - including custom GraphQL like feature.• Architected a multi-layered backend with dependency injection for mocking in Unit Tests using Moq and Xunit.

Fall 2018 - Current:• Lead a team of 8 to create a movie analysis tool based on the continuation of the project from Spring 2018. • Structured the multidisciplinary team into three sub teams that focuses in Computer Vision (CV), Natural Language Processing (NLP), and Audio Analysis. Spring 2018:• Created a retrieval based chatbot with personalities based on Disney movie characters using natural language processing (NLP) and cosine similarity.• The Chatbot was proposed to the Director and Manager of the Data & Analytics at Walt Disney Animation Studios, which successfully received continuation of the project for the year 2018-2019. • Hackathon: Implemented K-mean Clustering algorithm to compute the optimal coordinates for setting up emergency relief camp, in the event of an natural disaster, given users' critical data gathered by the chatbot. Received Best Domain Name Award.

• Worked on the development of an Android app that estimated the heart rate through time series facial imagery data. (Java)• Received $100,000 in funding from the Bill and Melinda Gates Foundation.• Translated kalman filter and signal processing algorithms from Python into static proxies using Chaquopy in order to implement the code into Java.

Command and Data Handling• Created the communication system in Python for Peroxide Engine Test Stand to send telemetry & command data between Ground Control Station and Guidance, Navigation, & Control subsystem using ZeroMQ.• Automated testing procedure by adding the flight profile and debugged the old UI of Ground Control Station.Both in C#.• Detected failure of the engine by implementing softkill which processed thermocouple, pressure transducer and load cell data readings using Python.• NASA PA Space Grant Scholarship.

• Provided troubleshooting of any computer problems the school encountered.• Replaced/fixed nonfunctional computer parts and software.