David Rendleman's Email Addresses & Phone Numbers

Software Engineer, Machine Learning @ Facebook Developed a real-time deep-learning based visual object detection system for Oculus Quest, Rift S, andother unreleased Oculus VR headsets.The system was responsible for semantic labeling of surfaces (walls, desks, ceilings, windows) as well asdetecting the 3D poses of objects (keyboards, mice, monitors, cups, office equipment, etc).Platform: C++, Python, PyTorch for Linux...

Software Engineer, Machine Learning @ Facebook Developed a real-time deep-learning based visual object detection system for Oculus Quest, Rift S, andother unreleased Oculus VR headsets.The system was responsible for semantic labeling of surfaces (walls, desks, ceilings, windows) as well asdetecting the 3D poses of objects (keyboards, mice, monitors, cups, office equipment, etc).Platform: C++, Python, PyTorch for Linux From June 2019 to August 2020 (1 year 3 months) Menlo Park, California, United StatesMachine Learning Software Engineer @ NVIDIA Developed, refined, and deployed a real-time deep-learning based visual object detection system in a civilian passenger vehicle targeting level 4 autonomy, as part of the deep-learning for autonomous vehicles team. Detection targets include cars, trucks, pedestrians, cyclists, signs, lanes, & drivable area.Engaged OEM customers directly, tailoring DriveNet and the Driveworks platform to suit manufacturer-specific ADAS and autonomous vehicle needs.Platform: C++ for Linux, Python, CUDA From August 2016 to May 2019 (2 years 10 months) Santa Clara, CaliforniaMachine Learning Software Engineer @ NVIDIA Researched technologies for and helped develop Nvidia's DriveNet, the object detection component of Driveworks - an "end-to-end machine learning platform for self-driving cars".Part of the deep learning applied research team, focusing on automotive applications.Platform: C++ for Linux, Python, CUDA From October 2015 to August 2016 (11 months) Santa Clara, CaliforniaSoftware Development Engineer @ Amazon Developed and maintained vision analytics solutions for automating Amazon fulfilment center activities. Applications included a vision analytics platform operating on tens of thousands of cameras and spanning multiple continents.Distributed real-time vision analytics on low-power edge devices accomplished through GPGPU based solutions.Platform: Java, C\C++ for Linux From December 2013 to October 2015 (1 year 11 months) Software Development Engineer @ Amazon Developed applications for an internal video content distribution system. Responsible for maintaining the health of a North-America wide surveillance camera network amidst massive expansion, including automated solutions for auditing camera health.Platform: Java, Ruby, Ruby on Rails From August 2013 to December 2013 (5 months) Student Researcher @ University of Pittsburgh Project designed to assist in physical rehabilitation therapy by making repetitive exercises fun for the patient and more informative for the physician. Uses Microsoft’s Kinect and the Natural User Interface for skeletal and visual data Worked directly with physicians and patients to assess performance and limitations of the software Platform: C++ for Windows From August 2011 to December 2012 (1 year 5 months) PittsburghStudent Researcher @ University of Pittsburgh Project aimed to construct a self-learning AI for the game of Stratego, chosen for its loose similarity to Chess but added complexity of a partially observable environment. Using self-play, the agent would improve its own performance over time, without requiring expert knowledge. Platform: C++ for Windows From January 2012 to May 2012 (5 months) PittsburghStudent Researcher @ University of Pittsburgh Goal was to develop a program which, given sensory inputs such as wheel encoder data, relative positioning data similar to output from a GPS, computer vision, and infrared object detection, would be able to accurately move about and interact with its environment. Utilized the WowWee Rovio mobile robotics platform, a small Wi-Fi enabled robot with a web interface. Performed light service on robots in need of maintenance and tuning. Platform: Perl, C / C++ for *nix From January 2011 to May 2011 (5 months) PittsburghStudent Researcher @ University of Pittsburgh Goal was to develop and measure the performance of heuristics for selectively throttling the power draw of a computer cluster in accordance with load. Assembled and maintained an experimental cluster of 6 computers, including performance and power recording hardware, and developed tools to automatically collect and tabulate data from this cluster Platform: Java, Perl for *nix From September 2010 to December 2010 (4 months) PittsburghStudent Researcher @ University of Pittsburgh Goal was to develop a program which, given a static snapshot of Wikipedia, developed the answer to an unprocessed human language question – ex.: “What did The Legal Tender Act of 1862 establish?”. Given this Wikipedia dataset, created a library of possible answers and the features that would describe a question that needed that answer. Context aware - could disambiguate poorly written English sentences when multiple answers applied to them Developed in Ruby, Python, C++ for *nix, Java, Perl From January 2010 to May 2010 (5 months) PittsburghInterim - Software Developer @ Princeton Satellite Systems Wrote programs to aid in satellite attitude determination and navigation. Visual feature detection using modified SIFT (Scale Invariant Feature Transform). Feature pattern matching using RANSAC (RANdom SAmple Consensus) and rigid body transformation determination. Platform: C for *nix/Cygwin From June 2008 to December 2009 (1 year 7 months) Princeton, New JerseySummer Research Assistant @ Princeton University Image analysis, feature recognition, and pattern matching. Used LabelMe database to train a machine learning algorithm in recognizing objects embedded in arbitrary scenes (image recognition) Platform: Matlab From June 2007 to September 2007 (4 months) PrincetonClerk @ Saint John Vianney Catholic Church Worked as an administrative office assistant and receptionist at large community-based Catholic church. Received hospital calls, scheduled appointments, and triaged emergency counseling services. Responsible for keeping office computers and local area network up-to-date, free of viruses, malware. From January 2003 to June 2006 (3 years 6 months) Walnut Creek