Staff data scientist involved in a number of projects such as: • Engineering production data science containerized applications with full reporting, logging, and monitoring at scale using Docker, Kubernetes, Cloudwatch, Grafana, as well as other AWS services for application and algorithm monitoring• Deploying cloud-native data pipelines (Kafka, Kinesis, Sagemaker, EKS) to ingest vehicle sensor data at scale and predict events such as tire leaks, high temperature anomalies, and low tread• Incorporating Natural Language Processing models to ingest, tag, and summarize competitive intelligence documents for use within the company for non-domain knowledge experts• Utilizing shipment and production data to build out forecasting models for RvCC• Building image classification models for production in potential consumer applications • Helping develop and integrate cloud architecture solutions for large-scale company use

• Consulted for Makrwatch to develop and build a machine learning codebase called Panoptica (panoptica.me) that predicts demographic information about age, gender, and country from YouTube channels and videos• Integrated newly mined video information from the YouTube API with previously mined YouTube channel and video data, coalescing this into various PostgreSQL tables• Processed YouTube video corpora of 650 GB of data with NLTK and latent semantic analysis, which fed forward into a multi-label regression model that outputs age/gender brackets• Predicted current marketing campaign demographics with 85% confidence

• Analyzed biological RNA modifications in context to determine mutation rate and stop rate of common reverse transcriptases for six common Watson-Crick methylations• Incorporated random forest classifier algorithms to infer effect of nearest neighbor nucleotides on mutation rate and stop rate of each modification• Classified RNA modifications from unmodified and sequencing artifacts in native deep sequencing to 95% accuracy using a boosted decision tree classifier

• Researched dynamic RNA modifications in both mRNA and tRNA using deep sequencing and bioinformatic tools• Contributed to bioinformatic analysis in order to formulate a tRNA-sequencing method that could precisely identify modification fractions at single nucleotide resolution in heavily modified RNA• Discovered new modifications in a well-studied system as a validation for tRNA-sequencing, and also applied this to validate sites of mRNA methylation• Investigated a nuclear and mitochondrial demethylase ALKBH1, the first enzyme shown to demethylate tRNA