I manage the development of models utilizing social media data alongside real world data (RWD).The team I manage ingests data from various sources, and has been responsible for more accurately defining the strategy for several of J&J's products. Our analyses tend to focus on the timing of patient need in their care journey, and providing an effective timing to prospective patients.In addition, we have explored consumer confidence models for current and past patients.

Ensuring data from and for external partners is of high and consistent quality, primarily for oncology research. EHR data analysis for ETL, using data science to enhance disparate cancer center data elements. Improving rates of clinical trial enrollment through discovery of eligible patients. Developed new software product to automatically monitor and improve data quality across multiple clinical sites. Performing and enabling analysis in TransMed's distributed network of oncology databases. Examples of this include: identification of patient cohorts for proposed clinical trials, creating predictive models of patients becoming available for clinical trials (i.e. predicting progression and/or treatment failure), patient volume forecasting and clinical trial digitization.Thought leader for tackling challenges in clinical data science and clinical decision support, such as developing novel presentations of data for CRO/Pharma bids, cancer progression/treatment failure predictive survival models, treatment pathway analysis leading to clinical trial enrollment, automated data quality metric collection and analysis.Technical skills include:- rapid prototyping of novel analyses and presentations- MSSQL and Hadoop to extract data- STATA and R for analysis- Graphical presentations in Draw.io and C#- Delivery of analyses and recommendations to clients in Word and Powerpoint

Developing new app clinical decision support algorithms and software, for oncology and intensive care.Pediatric intensive care research includes: ventilator-associated pneumonia risk calculation (more recently V-A conditions), live mortality risk scores, longitudinal HbA1c control in type 1 diabetics, bedside monitor streaming analytics, effect of oxygenation therapy on saturation, readmission risk prediction, census tract Oncology work includes: machine learning for genetic drivers and pathways of breast cancer, data science development to enhance oncology records from variably-populated EMRs.

As an honorary member of faculty, I maintained links between St George's and my other employers as well as pro-bono analysis and publication of genetic epidemiology studies.

Research into the factors leading to type 1 diabetes and celiac disease, using statistical and epidemiological analyses. Use of SNP data from the Illumina Immunochip on genetically predisposed subjects to look for genetic and gene-environment associations. On a separate project, analysis of T-cell reactions to analytes in a cow's milk/hydrolysed formula intervention study to find changes in the immune response prior to development of antibodies.Working in the transition of the Pediatric Epidemiology Center into the Big Data realm, utilizing Hadoop to curate and analyze data in this developing paradigm.

Research into the factors leading to type 1 diabetes and celiac disease, using statistical and epidemiological analyses. Use of SNP data from the Illumina Immunochip on genetically predisposed subjects to look for genetic and gene-environment associations. On a separate project, analysis of T-cell reactions to analytes in a cow's milk/hydrolysed formula intervention study to find changes in the immune response prior to development of antibodies.