Erik Schmidt Email & Phone Number

Los Gatos, California, United States

Los Gatos, Ca

Machine learning and recommender systems for personalization. Projects spanning homepage row positioning, billboard selection, category and talent ranking, and explainable recommendations. Driving algo for cross-functional projects spanning ML, design, product, and engineering.

Oakland, Ca

Developed machine learning systems for music mood understanding and helped launch a collection of personalized playlists focused on specific moods and activities. Led the algorithmic architectural design for recommendation features in Pandora’s on demand music platform, Pandora Premium. This architecture drives autoplay and the magic wand playlist extension tool.

Oakland, Ca

Led the playlist and algorithmic design of Thumbprint Radio, a hyper-personalized radio product that in its first year was used by over 25 million listeners and spun over 7.6 billion tracks. Contributed of numerous playlist recommendation strategies to the Pandora listening experience, both personalized and non-personalized, context and content-based. Developed of machine listening systems to automatically understand the musicological content of audio utilizing the incredible data collection known as the Music Genome Project. Work spanning machine learning, recommender systems, and digital signal processing.

Researcher in the Music and Entertainment Laboratory (MET-lab), a group led by Youngmoo E. Kim, an MIT Media Lab Alum. The MET-lab performs research in digital media technologies, especially in the areas of sound and music. Performed research in music emotion recognition spanning the areas of psychology, signal processing, and machine learning. Mentored Ph.D. students in multiple areas of music informatics: automatic mixing of multi-track audio, content-based prediction of musical influence, content-based modeling of musical expression, and musical robotics. Lead instructor for Drexel University Winter 2013 course on Machine Listening and Music Information Retrieval.

Ph.D. student in the Music and Entertainment Laboratory (MET-lab). Developed machine learning techniques for automatic prediction of time-varying musical emotion from audio, spanning Kalman-filtering approaches and conditional random fields. Developed deep learning approaches for extracting acoustic features specifically optimized for musical emotion. Developed and implemented DSP libraries for acoustic feature extraction from audio including loudness, timbre, and harmony.

Developed DSP algorithms for real-time, multi-rate audio processing for 32 simultaneous channels and assisted in FPGA hardware deployment. Maintained and deployed bug fixes in low-level (assembly) user-interface software. Developed hardware for digital mainboards. Performed SMT component level debug and repair of prototype hardware and system beta testing.

Phd, Electrical Engineering

Advisor: Youngmoo E. Kim, Associate Professor of Electrical and Computer Engineering Thesis: Modeling and Predicting Emotion in Music.

Ms, Electrical Engineering

Advisor: Youngmoo E. Kim, Associate Professor of Electrical and Computer Engineering

Bs, Electrical Engineering

Senior design team leader: developed and prototyped a multi-channel wireless electroencephalogram (EEG) system. Designed a compact.