• Manage multiple client relations• Develop automated solutions and infrastructure to meet business needs• Perform anomaly detection on aircraft telemetry data• Build interactive model dashboards• Model time-series component failure• Conduct data science trainings and workshops• Setup R and Python Integrations with Spark Clusters and S3 data storage

• Wrote and designed over 150 lessons for the data science curriculum• Lessons included web scraping, scikit-learn, TensorFlow and network analysis • Led lesson plan development as instructional staff grew from 5 to over 50• Trained and managed instructors implementing data science curriculum• Prototyped deep learning NLP web app to reroute ZenDesk support tickets

• Designed and built multi touch attribution pipeline using AWS• Ingested, parsed, and transformed 20GB of data per day to populate database• Developed strategy for people based attribution using multiple data sources• Coordinated with engineers and senior leadership to develop report automation• Built engagement index dashboard to monitor campaign performance

Designed, manage and implement a range of projects from audience segmentation to media optimization. Some of my projects have included: • Building custom classification algorithms to determine brand safe media acquisitions • Crafting audience segments using clustering algorithms and custom features • Developing social media listening tools to monitor ongoing trends • Mining and analyzing internal sql databases

Metis is a 12 week immersive project based data science bootcamp that provides training for a number of tools including Python, javascript, SQL, D3.js and Hadoop. In developing this skill set I have worked on a number of projects including:• Movie recommendation engine using OLS regression• Interactive visualization of MTA traffic using Pandas, QGIS and CartoDB• Analyzing whether the primaries are indicative of the general election• Toolsets for Struggling Readers: Here I used Natural Language Processing to develop a number of automated tools for struggling readers and their teachers or parents. These tools included ranking text reading difficulty, an automated glossary of potentially difficult words and a subject classifier using topic modeling. • Adaptive Tutor: Using real student data from DeltaMath.com, I created a recommendation engine using collaborative filtering and Bayesian statistics to develop a platform that can continually respond to student's progress, and assign them appropriate problems based on their success rates on previous problems.

While teaching at Pace High School, I incorporated real world applications into everyday lessons and was able to raise pass rates on the Regents by nearly 20%. In addition to my teaching responsibilities, I created the school's first chess club, and designed a year-long professional development framework for my colleagues. During this time I also:• Created and modified mathematical lessons according to student needs• Aggregated and analyzed student data using Microsoft excel• Designed programming explorations for the Ti-83 calculator• Built and maintained school chess club