Aaron Lee Email & Phone Number

New York, New York, Us

New York, New York, Us

- Optimized pipelines by decreasing table load times by 30% by building, deploying, and fine-tuning Airflow DAGs for efficient data ingestion from s3 to Snowflake.- Reduced table materialization by 7x at no extra cost to the warehouse by managing the ETL/ELT pipeline to the data warehouse while implementing dynamic Snowflake stage and Snowpipe generation with DBT macros for automatic schema inference.- Designed and implemented a new data pipeline, merging monitoring and product data functionalities and leveraging AWS (Kinesis, Lambda, Firehose, S3, etc.), Terraform, Airflow, Snowflake, and DBT to migrate away from legacy software.- Tracked customer journeys totalling up to $30 million in ARR to prevent churn, aid in upsells, and product usage data by supplying dashboards supporting cross-functional teams (product, customer success, marketing, etc.)

New York, New York, Us

- Managed ETL/ELT pipeline to the data warehouse (DAG, fivetran, stitch, upsolver, etc.)- Modeled, transformed, and validated data (marketing, GA, FB ads, and event-tracking, etc) with SQL/python and DBT/airflow.- Provided ad-hoc analysis and built dashboards supporting cross-functional teams (product, customer success, marketing, etc.)

New York, New York, Us

- Collaborated with Engineering and Sales teams to maintain, investigate, and debug critical data pipelines connecting to Sisense -- totaling $500k+ of business- Developed internal documentation in Confluence for the team and onboarded new hires- Maintained and built upon an application in Ruby on Rails connecting various API’s to query our Redshift Warehouse and deployed with Concourse -- creating a seamless workflow to access customer information- Consistent top performer -- Most tickets closed for 5 consecutive quarters Q3 2020 - Q3 2021- Consulted on live customer calls regarding data warehouse infrastructure options, Git capabilities, and database connection issues as the subject matter expert

Boulder, Colorado, Us

- Deep Learning on Images - Built an ensemble convolutional neural network trained on 50,000+ images to predict attraction and various machine learning algorithms (logistic regression, random forest, and neural networks) to feature engineer.- Optimized Models to Predict Doodle Facial Features - Achieved 95% accuracy in predicting facial features (100,000+ training images) using a range of classifier models: Logistic Regression, Decision Tree, Random Forest, Gradient Boosting, Support Vector Machine Classifier, and Grid Search.- Predict Fraud Detection Dashboard - Analyzed imbalanced dataset for eCommerce tickets to predict fraudulent event inquiries. Responsibilities include data exploration, cleaning, and predictive modeling using Gradient Boosting, K-nearest neighbor, Support Vector Machine, Grid Search, confusion matrices, etc. - Churn Predictor - Utilized gradient boosting and random forest classifier to analyze 50,000+ user accounts with 77% accuracy. Proposed 4 different promotional strategies based on profit matrixes and receiver operator characteristics (ROC) curve.- Movie Recommender System - Dealt with cold start issue using Surprise package’s KNN model along with content-based feature engineering and KMeans clustering to recommend similar movies.

San Francisco, Ca, Us

- Guided, taught, and assessed students across all primary school subjects- Strong presentation, communication, and instructional skills: Implemented slack and trained coworkers in my department on how to use in simple language- Adept problem-solving abilities: Noticed 5th grade students didn’t know how to type; researched and applied a school-wide typing program from 1st-8th grade- Designed school-wide promotional materials with Adobe Illustrator: calendars, seasonal programs, student handbook, etc. - Managed school-wide computers, printers, AV systems, etc.- Led corporate and church volunteers in 150+ hours of community service activities

Santa Cruz, Ca, Us

- Led classes empowering youth to engage with their communities surrounding issues of driving and drinking safety- Worked with 3 supervisors to coordinate Friday Night Live, a youth program 6+ sites across Santa Cruz County- Collaborated in a county-wide initiative with over 55 local establishments to impact drinking and driving statistics- Outreached to start 2 chapters through continuous contact with school administrators

Data Science Immersive

Bachelor’S Degree, Cognitive Science (Ai / Human Computer Interaction)