Alex S. Email & Phone Number

Hybrid-remote Workplace, US

Birmingham, AL, US

• Rebuilt the shop time estimation model from the ground up increasing accuracy by over 25% while simultaneously refactoring the API to consume historical features from a separate Airflow job and implementing a parallel.
• Continued mentoring and onboarding new data scientists and reviewing their work
• Created a new deployment framework and migration process for Data Science batch workflows with Airflow using a custom designed Airflow operator to streamline deployment, standardize logging and alerting, and decrease.
• Finalized and launched a new shared Python tooling set to use used across various data orgs bringing together multiple teams to help with maintenance, feature development, and knowledge sharing
• Led the retirement process of the DS AWS Batch clusters moving the workflows to Airflow and Kubernetes achieving the company wide tech goal of becoming cloud agnostic with standardization across multiple data teams

Birmingham, AL, US

• Designed, built, and deployed the shop time estimation model translating from R to Python to deploy the first real time model at Shipt in a Flask API via Kubernetes and allow for the launch of a new micro service to.
• Built a product entity extraction model with Spacy including a custom tagging module and built in parallelization to enhance search result accuracy
• Refactored the product brand name classifier model and added parallel processing to decrease computation time 5x, created a QA process, and designed a deployment strategy with the Catalog team
• Continued to maintain and develop all data science infrastructure, adding features like alerting, resource specification, message queue integration and base Docker images
• Developed an algorithm to create personalized autocomplete suggestions to be used by the Search team to tailor results to individual customers based on past search and purchase history
• Led the development of a shared Python library containing internal tooling for data processing, datastore manipulation, and AWS access

Birmingham, AL, US

• Worked on a team of 3 that designed and wrote the partner data ingestion framework, which processes 500+ million records daily including inventory, pricing, and images spanning over 100 partners. This allowed for the.
• Responsible for building and maintaining all infrastructure for the team (AWS Batch, Lambda, and RDS)
• Developed a custom CLI to manage and modify the AWS infrastructure and imports to allow for easy onboarding of new partners and imports decreasing turnaround time to under 2 weeks
• Built a local Docker workflow for testing imports that encompassed pulling sample data from production, setting up the Docker networking for a local Postgres test database, and running the services that sent data to.
• Responsible for designing a new interview exercise for candidates, phone screens, and interviews across two DE teams, along with on-boarding
• Mentored a number of team members helping them grow to senior roles

Birmingham, AL, US

• Developed a new infrastructure stack and deployment process utilizing AWS Batch and Lambda to run Docker images built with Jenkins/ Drone with GitHub web hooks to streamline the process for data scientists to run batch.
• Co-authored a Docker base image for the team to use across both R and Python for automated jobs
• Built a custom package for fuzzy string matching of product names translated from R
• Wrote a simple out of stock tracking model to prevent large swings in availability in the app due to incorrect data, which was implemented in the partner data ingestion framework across all data imports
• Monitored and fixed any failed nightly jobs as needed
• Brainstormed, tested, and helped debug models with the team

Birmingham, AL, US

While working for BBVA as a Data Management Specialist, I was one of the first members of the new Big Data team and eventually transitioned to it full time. We pioneered the new Big Data platform at the bank utilizing Confluent Kafka for creating data pipelines, Spark for analytical processing, and HDFS for distributed storage.As a proof of concept to.

Birmingham, AL, US

• Developed the data mapping specifications to integrate a new loan system's data into the bank’s core systems as part of the Data Management team.
• Assisted business users with defining business requirements and translated these requirements into technical documentation and logic for the ETL development team.
• Explored existing databases using complex SQL queries to refine the technical requirements and verify data accuracy and calculations.
• Produced data mapping specifications for the enterprise data warehouse and downstream data marts.
• Began the project to reverse engineer the customer profitability system including business calculations for inclusion in the EDW rebuilding project.

Birmingham, AL, US

• During my time as a graduate student in the Department of Physics at UAB, I studied materials for applications in alternative energy. My thesis project involved the creation of perovskite based materials for solid.
• Synthesized plots and tables from collected data to analyze results and determine trends.
• Mentored undergraduate students and supervised their work.

Birmingham, AL, US

• Taught introductory and advanced physics labs and classes to over 400 students
• Instructed physics majors in advanced labs data and error processing in Excel. Some topics covered were histograms, error propagation, and curve fitting.
• Built and overhauled a new teaching lab space for advanced labs, while implementing new experiments.
• Trained new teaching assistants and helped overhaul the lab curriculum.
• UAB Physics Department 2013 Teaching Assistant of the Year.

Bachelor’S Degree, Physics

Master’S Degree, Physics