• Develop AI models to benefit public health.• Standardize scalable methods for AI model deployment.• Work with clients to ensure business requirements are data driven.

PVAI: AI for pharmacovigilance (drug safety monitoring) for major pharma clients.• Led Data Engineering for data science & data engineering team, including data ETL, AI standardization and development.• Led AI discussions with pharma clients, and set requirements in agile development process.• Developed ML methods for model ensemble and orchestration.

Built data analysis pipelines for pharmaceutical industry applications using open source data and Python, including Pandas, Scikit-learn, NLTK.

Birch Forest Systems is a software, data science and technical marketing business. Projects include:• Backend web development for pluvio.com.• Predicted high & low risk loan customers using classifiers with cross-validation statistics.• Preliminary heart disease prediction from patient data.• Team Wingfoot, created marketing plan for mobile heart sensor for health hackathon.• Founders Hookup, Ruby on Rails web app to connect company founders.• Technical Writers, analyzed 300+ marketing documents for Agilent product rebranding.• Lead instructor at Railsbridge, SF for Ruby on Rails web workshops.

Taught Scripting Languages class (CMSC 433) for the CSEE Department, including Regexp, Bash Shell, Ruby, JavaScript, and PHP.

• Created sparse non-uniform nD data processing algorithm using multi-threaded Java with Python API, reducing data collection time 10x.• Wrote signal filtering and processing algorithms in Java with Python API for chemical structure analysis suite.

• Optimized DSP algorithm for new FPGA-based system, reducing hardware cost $2.2M annually. • Successfully defended Digital Signal Processing (DSP) patent in EU patent exam. • Led software triage team in Java / C / MessageQueue for high-throughput robotic sample handler, resulting in multi-customer sign-off and acceptance.• Project manager for Java / C data processing software upgrade, including external beta test. Software released Q1 2009.• Created data acquisition & processing software for $600k MR hardware systems for chemical structure and imaging applications, annual sales $140MM.

• Created, tested and patented DSP data collection algorithm, reducing signal distortions 10x. • Led iterative software development with engineers & chemists for experiment manager and automated analysis with robotic systems in Java/C, for unified user experience across product lines.• Implemented C backend for multiprocess data and parameter sharing with Java UI, for Solaris to PC migration.• Automated magnetic field optimization with gradient shimming by SVD, reducing calibration time 90% and increasing customer & manufacturing productivity.• Advanced customer support.