Led DynAgility’s Data Science team, implementing machine learning and product improvement strategies for Fortune 500 companies, and staff augmentation for Silicon Valley startups. Met with clients to access needs and present solutions, while leading internal data science research.Some highlight achievements include:► Architecting and building a COVID-19 mapping/projection and visualization tool, building the database and integrating data from the CDC and Johns Hopkins for the National Minority Quality Forum used by both the Trump and Biden administrations.► Conducting big data analysis for AOL, converting passive data sources into consolidated resource to actively identify, mediate and message user population, creating a model to identify reasons for churn while building an automated customer response program, significantly increasing customer retention.► Leading project to build and apply Latent Dirichlet Allocation (LDA) to government contracts, patents, and resumes for US government contractor, developing Topic Modeling to aid in novel search capabilities, significantly decreasing non-match rejections.► Applying pattern recognition and Natural Language techniques to create a program to perform name resolution on database of mixed personal and corporate names for real estate company, and performing Named Entity Recognition (NER) on material data safety sheets (MSDS) to automatically extract information.► Running feasibility study for advanced catalog searching with topic modeling (LDA) and image recognition (Keras and OpenCV) techniques for trade show and catalog company.

Machine learning and data mining analysis of genomic and clinical data for large hospital system in Northern Virginia, working cross-functionally with scientists, medical doctors, and clinicians to bring translational medicine to reality.Some highlight achievements include:► Spearheading Translational Medicine project, using genome and disease data to predict health problems, building a Laboratory Information Management system (LIM) by disparate information sources.► Establishing core architecture for managing lab and clinical information, integrating hospital electronic health records into data warehouse.► Evaluating technologies for supporting multi-terabyte genomic data.► Machine learning and data mining analysis of NGS (Complete Genomics) data and clinical data, and performing independent analysis to survey structural variations across multiple genomes.

Independent consultant supporting National Institutes of Health (NIH), NeoDiagnostix, Tessarae, and 5AM Solutions with Data Mining and Natural Language solutions.Some highlight achievements include:► Applying machine learning and natural language techniques to aid the NIH in their reporting of grants to Congress that were rewarded, while enabling them to select future grant winners based on historical data/language. Requirements support for caBIG Protocol Lifecycle Tracking System (PLTs).► Working with cervical cancer diagnostics start-up to enhance and support database and software changes for primary enterprise system for the build of genomic diagnostic tools.

Lead developer of four-person Informatics team, developing software and analysis for the discovery and optimization of small molecule compounds for the treatment of cancer. Produced cutting-edge bioinformatics research in pharmacology, gene transcription, and cellular pathways. Some highlight achievements include:► Bioinformatics Lead for Optimization Team on five major drug optimization programs, examining candidates for novelty and mechanism of action.► Biomarker Team member; Clinical Phase II Pancreatic/Phase I Leukemia cancer compound, AVN944, performing analysis on clinical data to select gene transcription markers that stratify patient populations and identify molecular characteristics of response and resistance to drug.► Developing novel software solutions that integrate domain-specific data into clear understandable products for Biology and Chemistry teams.► Architecting and developing graph tool enabling scientists to visualize metabolic network differences between active compounds saving years and millions of dollars on drugs that would fail development.► Developing a Java-based tool for examining gene transcription profiles for Microarray and Taqman platforms.

► Bioinformatics expert for Amplicon Confidence Value Algorithm, developing machine-learning solution to predict PCR success prior to lab validation, saving hundreds of thousands of dollars in lab costs to produce the VariantSEQr product.► Building a system to identify valid polymorphisms using machine-learning tools, reducing the required manual workload from fifteen human-years to three.► Inventing and patenting automatic classification software for identifying clusters within fluorometeric genotyping data, dramatically reducing human workload and increased throughput of genotyping pipeline.