Building, designing, and maintaining data models in table format by aggregating unstructured and structured data from disparate sources that will then be analyzed via ML algorithms for business process development and insight. Use of ETL products including Hadoop, Hue, Hive. SQoop, Kafka, MongoDB, Snowflake, AWS, GCP, Azure,etc. to build ETL streams that are then processed for algorithm efficiency decreasing query response times avoiding Big O notations. Data pathways experimentation to assess optimal streaming configurations for eventual deployment. Various memory reduction techniques that maps data to larger sets of information such as denormalizing of excessive tables to increase query efficiency.

Define table constraints and relationships utilizing information including ER diagrams, and personnel interviews to understand software architecture to translate into data results. Table aggregate calculations to extract specific metrics from large datasets(over 100 million rows) to determine meaningful insight. Responsible for extracting information to discover unique SKU identifiers per retail locations in multiple metro-statistical areas to determine quarterly expenses, profits, and revenue. Identified Cost of goods sold, days to receivables, days to payables which were then utilized to calculate cash to cash days via Teradata queries. Applied query results to determine marketing strategies (customer segmentation analysis) to increase quarterly profits year-over-year. Analysis of entity relational databases to determine cardinality constraints to detect anomalies in datasets so as to rule out information creating incorrect analysis. Creating of entity relational diagrams and schemas for better understanding of how tables relate and react with other tables within a database. Performed sentiment analysis utilizing Twitter API via data streaming pathways to ascertain multiple days of text information to develop marketing strategy. Ascertained analysis of text data results by applying python packages to provide subjectivity as well as positivity scores to determine validity of information. Resulting scores from analysis were then applied to customize a marketing strategy for client. Utilized Numpy, Pandas, Seaborn, SciPy, Scikit learn, Matplotlib in Python specifically to train cardiac enzyme models to determine which cardiac enzymes would be the best predictors of cardiac arrest. Implemented same dataset in H20 to verify validity of results as well as the variances in each programs results.

Train multivariate linear regression models to predict housing prices for a given MSA (metro-statistical area). Implementation of logistic regression algorithms to determine if microchips manufactured by a particular design company could pass Quality Assurance testing via decision boundary design. Implementation of an anomaly detection algorithm to uncover anomalous behaviour of network servers via gaussian distribution. The anomalous servers identified were then placed under intense statistical investigations to determine possible root cause(s). Practiced application of K-means clustering algorithms to compress file images which can be used to decrease storage capacity as well as minimize delivery latency to desired destination (-40ms). Principal Component Analysis (PCA) algorithm training was also used which was very useful in reducing the dimensionality of datasets that contained excessive or duplicate features which often can stall computational discovery due to asymptotic complexity. Gradient Boosting Models applied to aggregated and non-aggregated sports data to determine mock pay scales. GBM also applied to cardiac enzymes and EKG data to train models to predict key parameters to predict outcomes utilizing parameter tuning. Trained a logistic regression based neural network used to identify handwritten digits (0-9) by feeding 5000 20 x 20 pixel grayscale images to the model. Feature engineering was also employed using univariate and multivariate analysis (ie. chi square testing, heat maps)determining correlation between categorical and continuous data. Had the responsibility of developing Polynomial regression algorithms to determine net gain/loss of daily and monthly water levels of U.S. dams. Trained Support Vector Machine model to act as a email spam classifier employing 4000 different examples containing spam, and non-spam email.

Pharmacist duties but not limited to verification via Physician order entry. Implementation of Discern analytics via Cerner for experimental query reporting. Review of SAP reporting results to anticipate high demand, and low census periods to minimize revenue cycle expenses, and increase patient satisfaction.

Successful in decreasing delivery time of medications to nursing staff by an average of 10% within 2 months increasing overall patient satisfaction survey scores. Increased patient discharge consultation compliance scores by 6% over a 5 month period. Achieved clinical analysis of patient metrics to tailor medication dosages to reduce overall adverse effects and increase patient compliance.

Responsible for market analytics research via web scraping techniques to identify optimal restaurant establishments for revenue and profit building. Successful in matching Ph.D.s' in-house beer inventory with clients food menu via market analysis/linear regression modeling. Applied sentiment analysis to determine optimal potential clients best paired with the brewery services. Increased market penetration via customer service oriented skills building relationships with managers and restaurant owners. Responsible for securing 3 key accounts that lead to projected profits of 12% within first 8 months. Converted ingredients lists to cost effective alternatives via supply chain analysis which also allowed prediction of possible interruptions in beverage production and supply. Incorporated data analytics driven decision making (A/B testing) to determine best package designs for client products. Utilized Excel spreadsheet functions to populate data consisting of local suppliers of brewing ingredients. Utilized web scraping techniques via Python/scrapy to update ingredients cost, and applied results to linear regression models to predict future quarterly costs. Utilized profit and loss statements to project expenses along with potential revenue to guide growth.

Created graphic and mixed media designs for clients in various industries (food & beverage, entertainment, fashion , freelance) by applying linear regression models to completed questionnaires which assisted in determining optimal product design for clients. Responsible for target audience identification and creation of mediums to maximize exposure of respective goods and services. Analyzed similar service providers for insight to maximize client exposure and profitability. Accountable for revenue maximizing logistics of cost effect materials for various product designs and successful in identifying possible sources that could limit product penetration and marketability. Market design attributes applied lead to more than 10 clients within 3 months of initiation. Product design profits reduced expenditure costs by 16% in first year of operation.

Achieved efficient analysis of patient metrics as it relates to compounding Total Parenteral Nutrition requirements for clients' timely delivery. Data entry of patient metrics as well as physician treatment plans followed by delegation to pharmacy technicians to complete patient orders in addition to delivery requirements. Assisted management in increasing client satisfaction scores during the periods of 2011-2012 which lead to an 12% increase in new patient referrals by physicians and nursing agencies. Responsible for assisting management in curtailing costs of perpetual inventory below $200K , meanwhile providing efficient delivery of products and services to home infusion clientele. Instrumental in maintaining sterile compounding environment in CLASS 5,7, and 9 requirements providing supervision over technical staff of 8. Adhered to USP 797, USP 800 sterile guidelines with annual testing which identified particulate matter known to compromise a sterile environment.