Patrick Phelan Email & Phone Number

Denver, Colorado, Us

Building AI/ML infra

Pittsburgh, Pa, Us

▪ Hold office hours and guide students through topics such as Policy Gradient Methods (PPO, TRPO, NPG), Deterministic Policy Gradients (DDPG), Evolutionary Methods for Policy Search, and Model-Based RL. I also cover more foundational topics like Multi-Armed Bandits, Monte Carlo methods, Temporal Difference Learning, and exploration techniques such as Thompson Sampling and UCB

Seattle, Wa, Us

▪ Developed EC2 console widget, leveraging AWS services and CDK to architect a scalable backend infrastructure consisting of Lambda, DynamoDB, and API Gateway.▪ Applied two newly inherited frameworks, the widget framework and EC2 test console framework; and implemented tool to alert users of the statuses of their EC2 requests in real time; previously no support for notifying user of failed launches.▪ Assisted senior engineers by guiding the setup of new widget development environments; created documentation on communication protocols between the two systems to simplify future development processes.

Pittsburgh, Pa, Us

▪ Mentored a community of 200+ residents over 2 years; hosted events w/50+ residents in attendance

Pittsburgh, Pa, Us

▪ Developed automation tool to seamlessly merge and synchronize sensor data across the lab’s measurement devices: gas chromatograph, PTR-MS, gas concentration analyzer, and lab jack▪ Implemented a low-cost regression model (~$150) utilizing statistical software such as R, pandas, and Excel; achieved comparable performance to a high-end methane sensor worth $20,000.

Pittsburgh, Pa, Us

▪ Co-instructed a group of 30 students for CMU’s 15210 Sequential & Parallel Algorithms and Data Structures class through engaging weekly recitation sessions; conducted code reviews; provided personalized guidance to struggling students▪ Collaborated with course instructor and fellow TAs to ensure consistency in course delivery, align grading standards, and contribute to the overall enhancement of the learning experience▪Topics taught range from basic DP to graph contraction, MSTs, and randomized algorithms ▪Class taught entirely in SML

Warrendale, Pennsylvania, Us

▪ Evaluated 3D reconstruction software, Meshroom, for sewer system inspections, determining its ability to handle terabytes of data and generate panoramic images.▪ Replaced the need for inspections using thousands of 2D images with an interface akin to Google Maps' Street View by implementing Meshroom pipeline to assemble individual panoramic images into seamless 360-degree videos.▪ Suggested hardware updates to improve renders such as camera selection, rig calibration, field of view, camera position, camera-synchronization, and sensor size; provided pros and cons alongside each proposed update.

San Jose, California, Us

▪ Conducted analyses on direct-mail, tv, and email campaigns▪ Built SQL queries and integrated salesforce, postie, and dial800 data to create automated self-updating reports to track KPIs for CMO and VP of marketing▪ Developed new data-ecosystem and reporting tools in Google Data Studio to organize multiple data sources and facilitate weekly reporting▪ Conducted market research to evaluate upcoming tv-creatives; analyses aided in determining tv-creative direction

Palo Alto, California, Us

▪ Developed data-analytics program to manage and track driver performance▪ Streamlined data-processing system for billings department by automating weekly driver payment reports; reduced driver payment processing times from 2-3 hours to under 15 seconds

B.S. Artificial Intelligence, Minor In Statistics

Bachelor Of Science - Bs, Artificial Intelligence

Education record