Paul Mccabe Email and Phone Number

Spearheading logistics-optimized geospatial capabilities to optimize Amazon deliveries, including turn-by-turn navigation, sensor data processing, simulation and experimentation capabilities, and datasets to support machine learning.

Drove data collection, ML model evaluation and deployment, insights/analytics, and competitive benchmarking for Alexa speech recognition.• Pioneered the use of machine-generated datasets for evaluating and training speech-recognition models.✓ Decreased the time needed to recognize new and trending words from 3-6 weeks to less than one week.• Automated evaluation of on-device models using a physical device lab.✓ Reduced manual effort to evaluate models from 2-3 hours to minutes, increasing science velocity and enabling weekly production model updates.• Delivered an influential analysis of fairness in speech recognition.✓ Directly led to the creation of a new SVP-level program

Responsible for the ML platform and runtime services for knowledge-graph queries.• Integrated and scaled graph-building technology from an acquisition with Siri’s existing infrastructure and Apple-owned datasets.✓ Increased the size of the knowledge graph by a factor of 2.• Automated and optimized data pipeline reliability and speed.✓ Increased graph update frequency in production from 1-2 weeks to daily.• Coached and mentored engineers who came in through a startup acquisition.

Owned the routing, guidance, traffic, and vector maps services for Apple Maps• Hired expert routing researchers and modernized routing to use the state-of-the-art algorithm.✓ Improved route quality by 10% and reduced latency by 50%.• Pioneered the use of machine learning for time estimation.✓ Improved the quality of historical and real-time traffic data and ETA accuracy.• Launched user-delighting features such as public transit and lane guidance.✓ Increased customer satisfaction and closed competitive parity gaps.

Developed key components of the technology stack for real-time traffic.

Led a team of software engineers developing the routing engine for Nokia Maps. I grew the team from three engineers to ten and started development of a modern routing algorithm using contraction hierarchies. During this time we implemented traffic-aware routing using live and historical speeds, and a simple frequency-based transit routing algorithm.

In this role I helped to develop the map rendering engine for the Nokia Maps mobile application. Some of my most visible accomplishments include 3D landmarks, label placement and collision detection, support for various map projections, tile maps for the web, development of a fast 3D graphics pipeline, and driving development of OpenGL support. This was a great opportunity to learn about computer graphics and GIS software development.

I was a C++ software developer at a start-up that developed a visual internet chatroom, using Flash for the front-end, C++ and Java for the back-end, and CORBA for communication. I worked on the back-end servers and had the opportunity to learn a ton of useful software engineering skills while applying some computational geometry and graph algorithms.

During the course of my Ph.D. studies I taught "Introduction to the Theory of Computation", "Discrete Mathematics", and "Computational Complexity and Computability". I'm pleased to say that in spite of the difficult subject matter (which I love, but my students mostly hated) I had very good reviews from my students. Teaching was a great chance to develop my presentation skills and to practice explaining difficult concepts to disinterested audiences.