Hello, thank you for visiting.

I'm Allester. 🤝

Pleasure to meet you.

Thank you for visting. Scroll down to learn more.

    • About Me

    Thank you for taking the time to visit my portfolio. As a technologist, my interests are in software development, machine learning, and computer vision.

    I currently attend the Dual Masters degree program at Penn Engineering. So far I have graduated from the MCIT program and on track to graduate with MSE-AI in Spring 2027.

    After a summmer internship at Barracuda Networks as an Analytics Engineer over summer, I am doing another internship as a Software Development Engineer at Amazon, where I will work with the One Grocery team in AI and Machine Learning.

    Work

    Work-related experiences — incl. employment, self-employment, and side-projects.

    Data Scientist @ Vintra, Inc. (acq. Alarm.com)

      • Built the Cloud Product User Activity Dashboard (Tableau) to display user activity on the cloud product; created an automated ETL pipeline (Google Apps Scripts) utilizing Vintra’s API to access the cloud product database (MongoDB)
      • Innovated the Density Mapping Tool to dynamically change Tableau backgrounds based on applied aggregations through the development of non-OOTB features in Tableau (Google Cloud Storage)
      • Built and designed the North Star Metric KPI’s Dashboard (Tableau) for the internal product team and an end-to-end batch processing pipeline to handle processes from file ingestion to refreshing the dashboard’s data.
      • Developed a Density Mapping Tool utilizing object detection data from Vintra’s REST API; presented as a product demo to the VP of internal security at a large social media company with our head of sales.
    TableauGoogle Apps ScriptMongoDBSQLPythonGoogle Cloud Platform

    JUN 2021 - DEC 2022

    Data Science Fellow @ CCBER

      • Built a Photogrammetry Pipeline (Metashape) for reconstructing 3D bee models to generate comprehensive bee trait and image datasets; aided in the biological study of image and anatomical trait digitization of bees (Big Bee Project)
      • Reduced Reprojection Error (pix) from 0.84 to 0.41 and processing time by 75% of pre-existing bee models through improving data collection, data augmentation, and parameter tuning on 16+ photo sets containing 145+ macro images
      • Trained an object detection model (YOLOv8) on 8000+ labeled bee image data from 3D models and iNaturalist to recognize video frames featuring bees to reduce video analysis and image extraction time from hours to minutes.
    MetashapeYOLOv8Python

    Spring 2023

    View Full Resume

    Get In Touch

    Want to work together, connect or just say hi? My inbox is open, and I'll do my best to get back to you!
    Say hello