1. http://rachithrr.com Rachith R Ramaswamy +1 716 (986) 1569
github.com/rachithrr linkedin.com/in/rachith-ramaswamy rachithr@buffalo.edu
EDUCATION
University at Buffalo, New York Master of Science in Data Science (GPA: 4/4) August 2019 – December 2020
• Graduate Coursework: Artificial Intelligence, Computer Vision, Deep Learning, Data Models and Query Languages,
Bayesian Machine Learning, Supervised & Unsupervised Learning, Probability, Linear Algebra.
• Bachelor’s Coursework: Data Structures, OS, Algorithms, Image Processing, Programming in C, C++, Java.
WORK EXPERIENCE
Software Development Intern Breathing.ai May 2021 – Current
• Developing an application to adapt the browser theme and color based on the detected stress of the user and suggest
relaxing exercises. [JavaScript, python, AWS].
Data Science Intern Editorialist YX September 2020- December 2020
• Developed a personalized home-page experience on the e-commerce website using item-based filtering. Designed the
full data pipeline, pre-processing more than 5 million products, extracting client preferences and activity on the site,
and provide top 50 relevant products. [unsupervised learning, recommender system, Information Retrieval, AWS]
• Built the “relevant” sort order on search results to improve customer click rate on the first page of the search results
using hybrid content-collaborative filtering model. [PostgreSQL, machine learning, sklearn, nltk]
Research Assistant University at Buffalo October 2019 – September 2020
• Developed a data pipeline from collection to qualitative analysis of “Problem Typology” based on intramural projects
and interventions. Research funded by the National Science Foundation [Clustering, Python, R]
Software Engineer, Machine Learning Motorola Mobility LLC January 2016 – July 2019
• Built a “Context Awareness Engine” to track phone activity and usage patterns and identify activities like driving,
sleeping, running, at_home, at_work, etc. using on-device machine learning. This library-as a service was used by
7 other applications to serve their customers [Supervised learning, Clustering, TensorFlow, Java, Android]
• Developed a recommender system to predict the top 5 applications the user might use, and the settings on the mobile
phone using the phone usage pattern [Recommender system, Bayesian Statistics, Android, Kotlin]
• Built AT&T cross-platform messaging service with real-time syncing, cloud backup, and restore for Motorola phones.
Further optimized to run on low memory and low power devices. [ RestAPI, Java, SQLite, Android]
• Developed “MotoBody”, a fitness application for smartwatches to track steps, calorie-burn, heart-rate etc. analyse the
trends and provide insights on the mobile phone. This application was downloaded by over 1 million people on the
android play store with over 350,000 active users. [Wearables, Java, Android]
SELECT PROJECTS
Video Chat application using aiortc: Developed a peer to peer video chat application using aiortc, a python wrapper for
WebRTC protocol. [JavaScript, python, aiortc, multi-processing, NumPy]
Winner of MTA Back on track (Microsoft Hackathon): Performed feature engineering and statistical analysis on the
passenger data and built a model to predict the number of people traveling on a route on any given day using Azure services.
Performed hyperparameter tuning and improved the accuracy to 70%. [Azure Machine Learning, supervised learning]
Automatic detection of Graphene Layers: Built a convolutional neural network model to identify graphene in the material.
Utilized cascaded model to reduce the False-Positive rate and obtain an F1-score of 0.72[TensorFlow, Deep Learning, CNN]
News sentiment analysis to predict stock price: Built a combination of time-series LSTM and NLP module to monitor
news and predict corresponding stock price. [Recurrent Neural Network, LSTM, Keras, Natural Language Processing]
Fraud detection: Analysed performances of XGBoost and Light-GBM on a credit card dataset from Kaggle and obtained
a precision and recall of 0.75 and 0.60 after hyperparameter tuning [R, exploratory data analysis, seaborn, sklearn]
Face-Mask detection: Built an application to stream video from camera and alert when a person without a face-mask is
detected, using the face-mask dataset from Kaggle. [OpenCV, TensorFlow, Keras, Convolutional neural networks, Android]
TECHNICAL SKILLS
Languages: Python (3 years), Java (4 years), C++ (6 months), R (1.5 years), Kotlin (1 year), MATLAB (3 years), Scala.
Development: Android, Django, Flask, React JS, Maven, Gradle.
Technologies: TensorFlow, Pytorch, MySQL, MongoDB, Azure, AWS, Docker, Jenkins, Git.