work

I'm obsessed with tackling problems through applied AI, software development, and robotics. Here's an overview of my professional journey and key projects.

Boston Dynamics

Staff Machine Learning Engineer, Atlas Applications Software | Sep 2026 — Present

Amazon Robotics

Software Development Engineer, Robotic Sortation Technology - ML / Optimization | Dec 2024 — Sep 2026

Leading real-time ML and optimization systems for robotic package sortation at scale:

  • Designed, built, and led growth of an LLM-powered AI agent framework to monitor key package processing metrics, perform optimizations, and fix configuration issues completely autonomously within robotic fulfillment centers.
  • Led implementation of a real-time predictive optimization framework to predict and reduce internal package recirculation inefficiencies. Delivered in first six months with millions in projected annual savings.
  • Implemented new data fetching strategy for fulfillment metrics dashboards, reducing feature development time by 60%.
  • Developed microservices in Go, Kotlin, and Scala on serverless AWS infrastructure with strict monitoring, testing, and uptime standards across global fulfillment centers.

Loupt

Technical Co-Founder, 2023 — Present

Building a Regulation Crowdfunding platform to democratize access to investment opportunities:

  • Designed and deployed Node.js services and Next.js web application with PostgreSQL on ECS using AWS services.
  • Integrated third-party APIs for KYC/AML, payment processing, and marketing automation.
  • Led development team and collaborated closely with institutional investors and stakeholders.
  • Held regular Scrum meetings and enforced coding standards to maintain velocity and quality.
  • Still spend 90% of my day in the IDE — building, shipping, and iterating.

Amazon Robotics

Software Engineer Intern, QA Automation | Summer 2023

Built a serverless React dashboard for QA regression tracking:

  • Developed serverless React app using AWS CDK (Infrastructure as Code) for CI/CD and deployment.
  • Completed project weeks ahead of schedule and iterated based on end-user feedback.

Structural Steel Technologies

Software Engineer | Dec 2021 — Jun 2023

Led development of device interaction software for AI-powered construction assistive system:

  • 10×-ed wireless image transmission throughput using SIMD for real-time compression.
  • Built C++ interfaces for CV cameras, SBG Systems IMU/GPS, and Micro-Epsilon LiDAR.
  • Enabled centimeter-accurate GPS tracking using RTK positioning.

projects

CV Hazard Detection Model

Python, OpenCV, Keras, AWS SageMaker | Lehigh Capstone - Fall 2023

Fine-tuned object detection for autonomous cyclist safety with Bethlehem startup, Skillion:

  • Improved off-the-shelf model performance by 100x for cyclist hazard detection.
  • Trained on KITTI, NTNU datasets, and custom Tesla Model S onboard camera data (labeled in-house).

Newton Notes

SwiftUI | gasek.net/newtonnotes | 2024

A minimalist, privacy-first gym workout tracker built in pure SwiftUI:

  • Open-source alternative to bloated fitness apps — no accounts, no tracking.
  • Clean, intuitive UI focused on speed and simplicity.