Software engineer, audio analyst and data scientist with over five years of programming and DSP experience. I develop full stack high performance machine learning algorithms for data analysis, and machine improvisation (generative models) with formal specifications. – Machine Learning, Data Analysis, Music Information Retrieval – Python, R, Matlab, C/C++, SQL, Postgres, Node, Ruby... – Git, SVN, Agile

Researcher @ TerraSwarm @ Audio-based anomaly detection Design and implement formal specifications for control improvisation systems Provide Music Information Retrieval resources and frameworks – HMMs Factor Oracles Python Java Formal Methods From January 2014 to Present (1 year 10 months) UC BerkeleyPhD Student in Machine Listening and Improvisation (Interdisciplinary PhD) @ PhD Research Classification models for noise-like sounds from bowed chordophones Representations of noise flows (statistical summaries and temporal evolution profile) Incorporation of noise flows into the specifications of an Audio Oracle based improvisation agent Multi-objective search of time series – Dirichlet Variable Length Markov Models GLMs GMMs Parallel Computing From August 2012 to Present (3 years 3 months) 1750 Arch StreetData Scientist @ Develop algorithms for beat extraction, local and global key estimation and chord transcription. From June 2015 to September 2015 (4 months) Laboratory for the Recognition and Organization of Speech and Audio (LabROSA)Data Scientist @ Store occupancy prediction – Time series Analysis EDA ARIMA STL R Python http://blogs.wsj.com/venturecapital/2015/01/12/percolata-emerges-from-stealth-with-5m-for-new-sensors-for-retail-stores/ From November 2014 to February 2015 (4 months) Audio Analyst and Data Scientist @ Designed from scratch a machine listening engine to estimate room occupancy and activity from audio Increased room occupancy estimation accuracy by designing a sensor fusion (audio, video, wifi) algorithm Provided error correction algorithms for vision occupancy estimation – GLM GMM HMM Python Sklearn Stats-models librosa Matplotlib From May 2014 to August 2014 (4 months) Software Engineer @ Development and Maintenance of externals and patches for CNMAT’s MMJ Depot - (C, Javascript and Max MSP Jitter, SQLITE) From May 2013 to March 2014 (11 months) Berkeley, CABackend developer @ PIC Your Future Backend development, including computer vision algorithms . Basic server setup and management (OpenCV, C++, Python, Ruby, Javascript, HTML5) From 2013 to 2014 (1 year) BerkeleyElectroacoustic Director and Musical Assistant @ Lead and hire a team comprised of international musicians, actors and audio engineers . Develop multimedia software for real-time processing in the opera Berio Sem Censura. (C, Javascript and Max MSP Jitter). From January 2012 to August 2012 (8 months) E-Studio Tutor @ Instruct and supervise graduate and undergraduate students in the diverse domains of computer music, including computer assisted composition (OpenMusic), sound analysis and synthesis (AudioSculpt) and Max MSP Jitter, From January 2011 to July 2011 (7 months) Owner @ Rehearsing and Recording Studio. From 2002 to January 2008 (6 years) Rio de Janeiro Area, Brazil

PhD in Machine Listening and Improvisation, Machine Listening and Improvisation, 3.95 @ University of California, Berkeley From 2013 to 2017 Master, Computer Music / Composition, ECU and MH-Stuttgart @ MH-Stuttgart From 2009 to 2011 Rafael Valle is skilled in: Machine Learning, Python, Artificial Intelligence, Programming, HMI design, Digital Signal..., Algorithms, MySQL, Matlab, Composition, Music, Conducting, Lecturing, University Teaching, Teaching, PHP, Songwriting, Orchestration, Higher Education, C++, Java, C, JavaScript