Michael Steiner Email & Phone Number

Beaverton, Or, Us

My core day-to-day tasks in this role included:• Leading the research and testing of tools and infrastructure for seamless cross-team collaboration.• Establishing credibility and gaining buy-in from key stakeholders by delivering high-value deliverables.• Designing and implementing an end-to-end risk identification and mitigation pipeline.• Creating a predictive model and segmentation index to detect policy abuses.• Developing innovative tools and methods to identify scams.A section of my key accomplishments in this capacity entails:• Leveraging knowledge, tools, and insights from other Nike teams to maximize efficiency and effectiveness.• Training and equipping team members with the necessary skills to deploy and maintain data science solutions.• Developing and executing a policy abuse propensity scoring model using XGBoost, achieving a test F1-score of 0.82.• Completing a Proof of Concept (POC) for detecting shipping-based scams using supervised machine learning, resulting in a test AUC of 0.88.• Successfully delivering an account fraud detection tool that reduced investigation time by over 40% and decreased the false positive rate by 22%.• Implementing an online risk identification and mitigation pipeline, capturing, analyzing, and reporting on over 1,000,000 risk events since January 2023.Skills:• Data Science• Cross Team Collaboration• Stakeholder Engagement• Risk Identification• Policy Abuse Detection• Predictive Modeling• Segmentation• Scam Identification• Team Training• Model Development• Machine Learning• Fraud Detection• Data Analysis• Analytic Reporting• Data Science Pipeline Implementation

San Francisco, Ca, Us

My core day-to-day tasks in this role included:• Leading a team of Data Scientists and a Data Engineer in SAAS solution development.• Building, testing, and improving ML/AI models.• Managing and mentoring a diverse team of Data Scientists, from junior to lead, and supervising interns.• Conducting research and development of new models, algorithms, and techniques.• Overseeing ML Engineering and DevOps engineers and assigning tasks from the Data Science team.• Supporting and contributing to the integration of deep reinforcement learning into NBA SaaS.• Developing an end-to-end schedule adherence predictive pipeline.• Collaborating with the Product team to design and develop new features for the NBA SaaS and consulting projects.A section of my key accomplishments in this capacity entails:• Recruiting and training a team of six Data Scientists and providing guidance and mentorship.• Enhancing model quality metrics by at least 5% through the creation of an automated data drift detection tool.• Boosting quarterly suggestion acceptance rate by 10% through end-to-end schedule adherence predictive pipeline development.• Designing and implementing innovative sales prediction and constrained optimization solutions, resulting in an 11% increase in sales.• Leading the project to construct feature stores, reducing model development time by over 50% and improving technical communication within teams.• Spearheading the development of a Next Best Action (NBA) SaaS using a range of models, including Random Forest, XGBoost, Logistic Regression, and Neural Networks.Skills:• Team Leadership• Data Science• ML/AI Model Development• Mentorship• Research and Development• Project Management• ML Engineering• DevOps• Deep Reinforcement Learning• Schedule Adherence Prediction• Collaboration• Feature Engineering• Sales Prediction• AWS Management• Technical Communication• SAAS Development• Model Building

San Francisco, Ca, Us

My core day-to-day tasks in this role included:• Developing Python, R, and Java models as the foundation for simulations.• Acting as Scrum Master for the IT/operations team.• Collaborating with cross-functional teams to define business problems, identify data sources, and design solutions.• Staying updated with industry trends, research, and advancements in machine learning and data science techniques.A section of my key accomplishments in this capacity entails:• Enhancing the Java platform with ongoing support and improvements.• Developing an advanced AI system for multi-user, repeated cooperative game theory simulations.• Utilizing mathematical optimization techniques like Simulated Annealing and Hill climbing to improve performance.• Successfully completing various major projects, including: ◦ Writing diverse models covering subjects such as game theory, business analysis, group dynamics, and difficult negotiations. ◦ Contributing to the development of a new RESTful API using Java, Spring, and Morphia. ◦ Creating monitoring and support tools for AWS servers using Python and Boto3. ◦ Managing 8 projects using Agile Scrum methodology, assuming the roles of Scrum Master and data scientist. ◦ Implementing Epicenter API in Java with Hibernate and Morphia. ◦ Maintaining AWS environments (development, UAT, and production) with Ansible, Jenkins, and Boto3. ◦ Deploying server health monitoring and alerting systems like Datadog, Nagios, M/Monit, and PagerDuty.Skills:• Python Modeling• R Modeling• Java Modeling• Scrum Master• Cross-functional Collaboration• Problem Definition• Data Source Identification• Solution Design• Industry Trend Monitoring• Machine Learning• Simulation Development• AI System Development• Mathematical Optimization• RESTful API Development• AWS Management• Agile Scrum Methodology• Server Monitoring and Alerting• Data Analysis

Scientific Holdings LLC is a unique IP licensing and technology consulting firm with strengths in chemistry and intellectual property management. Built a Data Science pipeline that ingests chemical assay data and emits the biological class.Technical challenges.• Noisy data.• Data needed to be scaled because multiple types of lab equipment were employed.• Biological classes were not mapped to actual data.A section of my key accomplishments in this capacity entails:• Random Forests model for class prediction• DBCAN model for clustering chemical assay data into “natural” classes. • Using SME was able to create a (inverse) map from algorithmically found clusters to biological classSkills.• Segmentation• Supervised machine learning• Unsupervised machine learning

My core day-to-day tasks in this role included:• Designing and implementing Predictive Analytic Software tailored to CallSocket's specific requirements.• Programming machine learning software using the R programming language.• Utilizing expert skills in Talend Open Studio, MySQL, PostgreSQL, R, Excel, and Tableau to manage Talend jobs processing data from various sources.• Managing 8 offshore developers to achieve optimum outcomes.A section of my key accomplishments in this capacity entails:• Constructing and maintaining ETL pipelines for fifteen clients.• Significantly enhancing and streamlining the invoicing process.• Facilitating the integration of the QA/QC team into the reporting analytics process.• Conducting thorough testing of machine learning software and exploring alternative models for optimal performance.• Developing a lead scoring model resulting in a 20% increase in cold lead conversion rate and a 30% reduction in lead costs.• Initiating and overseeing the integration and training of QA/QC team members into the analytics process, leading to a 35% decrease in contested sales.Skills:• Report Generation• Predictive Analytics• R Programming• Talend Open Studio• Database Management (MySQL, PostgreSQL)• Data Processing• Data Analysis (Excel, Tableau)• Offshore Team Management• ETL Pipeline Construction• Process Optimization• QA/QC Integration• Machine Learning• Lead Scoring• Training and Integration• Cost Reduction• Client Relationship Management

Responsibilities and Contributions include:• Insuring the ongoing functionality and security of IT infrastructure and seamless operation of hospital facilities• Provided 27/7/365 IT needs of hospital staff of 80 employees• Responsible for hiring and training in-house IT team• Administer Active Directory, Backup Exec, Configure and administer Cisco switches, firewall and phone system. • Maintained ultrasound and digital radiology equipment and data repositories. Supported all equipment for in-house lab• Provided desktop support for Window 7 workstations• Administered a wide range of software, tools, and operating systems to include (but not limited to): MIP Sage Fund Accounting, Blackbaud Razor’s Edge Database, Impromed Infinity, Backup Exec, Microsoft Exchange, efilm, Proginy & eSeries

This paid research position focused on computational topology methods as well as traditional statistical methods to examine gene expression data. The objective was to explore how gene expression data and copy number data could be used for differentiating breast cancer subtypes as well as examining patient survival for each of the cancer subtypes. The primary methods and tools employed: R Programming Language, Concom (computational homology software), Matlab, Mathematica, and ANOVA.

M.A., Mathematics

B.S., Applied Mathematics

B.S., Physics

Non-Degree Program, Comprehensive Cancer Research Training Program

Non-Degree Program, Algebraic Topology