Email: m@mjschock | LinkedIn: in/mjschock | Phone: +1 925 878 5408

Phaidra | Seattle, WA (Remote) | Oct. 2022 - Jan. 2024

• Spearheaded orchestration and automation of AI agent training (with each agent an ensemble of PyTorch models) into an MLOps pipeline using Prefect Server and Agent, demonstrating scalability using the Prefect-Ray integration and Ray.
• Modernized the developer experience by implementing Tilt for Kubernetes manifest changes, enabling rapid iteration with custom functionality for running ad-hoc data preparation, agent training, and inference pipelines.
• Pioneered a proof-of-concept for cost-efficient scaling using SkyPilot to abstract Ray and cloud computing resources, optimizing for minimal computational cost and time.

Georgia Institute of Technology | Atlanta, GA (Part-Time; Remote) | Aug. 2022 - Dec. 2022

• Served as a Teaching Assistant (TA) for CS 7639: Cyber-Physical Systems Design & Analysis.

Greyscale AI | San Carlos, CA | Oct. 2021 - Jul. 2022

• Developed an end-to-end data engineering pipeline to process diverse image sources into the COCO dataset format with k-fold train-validation-test splits using the FiftyOne and Albumentations libraries.
• Architected and implemented a data modeling pipeline for training PyTorch Faster R-CNN models for computer vision tasks, incorporating advanced object detection and image segmentation capabilities.
• Created an automated model deployment pipeline using Docker and TorchServe/SageMaker, including scheduled model monitoring and performance visualization through Amazon QuickSight.
• Unified the data engineering, modeling, and deployment pipelines using Kedro and DVC, implementing intelligent artifact tracking for efficient pipeline execution.

Ople.AI | San Mateo, CA | Sep. 2018 - Oct. 2021

• Refactored the data ingestion pipeline into modular components, improving system maintainability and scalability.
• Led the development and implementation of the forecasting service using Amazon Forecast, enhancing predictive capabilities.
• Architected and built a distributed worker service operating on graph structures for complex machine learning tasks and state management.
• Drove the implementation of model explainability features using SHAP (SHapley Additive exPlanations).

Georgia Institute of Technology | Atlanta, GA | Jan. 2018 - May 2025

University Of California, Berkeley | Berkeley, CA