Andrew O.
About Me
I am an enthusiast who provides actionable insight into data and perspectives, which helps answer relevant questions relating to any initiatives (rewarded or unrewarded) to gain a competitive advantage and support effective decision-making.
My domain background is IT infrastructure and security, but I have some abilities spanning any full life-cycle quantitative initiatives as showcased by projects in my portfolio. My work is progressively being uploading to GitHub to demonstrate these capabilities, so please check it out.
GitHub – https://github.com/andyogah
Education
Harvard University (Extension School), Cambridge MA
Capstone Collaborative Project: Statistical modeling of a highly scalable predictive algorithm to save the classification time of magnetic resonance imaging (MRI) scans by 30% year over year and enhance the efficiency and accuracy of radiologists in the healthcare field Relevant Coursework: Database Systems (Computer Science), Statistical Methods (Mathematics), Introduction to Quantitative Methods (Economics), IT Finance and Communications.
Springboard, USA
Capstone Project 1: Classification model predicting terrorist attacks perpetrated by domestic or foreign nationals in the US using the Global Terrorism Database (GTD) and Random Forest algorithm in Python. Capstone Project 2: Testing multiple predictive models for fraud detection in the public health field using large datasets of over 25 million records from the Centers for Medicare and Medicaid Services (CMS) and semi-supervised learning algorithms. Associated Skills: data visualization, data mining, prescriptive & predictive analytics, machine learning, and big data analytics
Work & Experience
Analytics Consulting & Personal Projects
Freelance
1. Model selection of the SVM algorithm from various predictive cross-validated classification techniques using independent judgment to research data science solutions and the F1 score and AUC/ROC performance metrics results for a client on Upwork. 2. Predictive classification model to ascertain the likelihood of a renewal of policy premium collection with an added incentive plan to maximize the policies' net revenues in the Mckinsey & Co. / Analytics Vidhya Hackathon. 3. Creative feature engineering to improve prediction of patient wait times from the Massachusetts General Hospital's operational data. 4. Feature selection of the best predictors for customer retention on a JSON time series dataset in the Ultimate data science challenge. 5. K-Means clustering for customer segmentation with the Elbow Method to research the optimum number of clusters to aid the development of customized marketing campaigns for a client on Upwork.
Technical Analyst.
Charter Communications (Spectrum)
1. Provided support services to top-level clients, creating value for products, ensuring more than 90% client retention. 2. Collaborate with the Regional Business Operations Center to troubleshoot and resolve large-scale technical issues using internal tools and documented procedures, with the maintenance of 99% availability of service. 3. Assisting in reducing service support shrinkage by 10% by working in cross-functional teams to impact customer experience positively 4. Involved in the calibration and documentation of baseline metrics for behavioral quality assurance control in speech and voice analytics to maintain frontline support efficiency at 75%, enhancing continuous improvement in workflows.
