ADAM KULIKOWSKI

Developed a React-based map dashboard in Cat Central to generate heatmaps from machine GPS (latitude/longitude) data, analyzing spatial patterns to optimize site layout, reduce congestion, and monitor machine idle-time, integrating real-time telemetry and secure machine communication for data-driven site optimization..

Developed RESTful API specifications and proxies using Swagger and Apigee, integrating JSON threat limiting, threat protection, rate limiting, and injection prevention policies for large data applications & real-time asset monitoring.

Built an AI-assisted CI/CD pipeline (Python, GitHub Actions, Azure Foundry) with a custom OpenAPI linter and automated proxy validation, reducing spec review time by ~30% and enforcing standards across engineering teams.

Selected from 70+ interns as winner of Caterpillar U.S. Hackathon to represent Cat USA at Global Hackathon.

Engineered data aggregation and labeling to train "DolphinGemma", an AI model designed to mimic wild dolphins.

Processed over 3,500+ bioacoustic audio samples (Python, Audacity) to prepare training data for DolphinGemma, ensuring high-quality, structured input for downstream machine learning development.

Featured at Google I/O and highlighted on the Google homepage as a showcase of applied AI.

Developed a real-time fencing broadcast animation system for tracking the movement of fencing blades for referring & TV.

Used OptiTrack infrared camera input, Python OpenCV for motion tracking & a dynamically tuned Gaussian blur parameter.

Finalist on PBS's Inventure Prize in front of 60,000+ live viewers on TV, winner of Georgia Tech's Create-X program.

Spearheaded $20,000 financing deal for the software to be used in an upcoming fencing related TV production.

Designed and implemented Spotter's Guide, an iOS catalog app (SwiftUI, Xcode) for Caterpillar machines, engines, & parts.

Structured JSON machinery datasets, created UI/UX, and published to the App Store (TestFlight, Apple Developer Tools).