David Schwab Email & Phone Number

Indianapolis

Developing and maintaining the main portal for local officials and other stakeholders to access budget and other fiscal data. This portal serves over 10,000 users and is crucial for ensuring on-time tax billing for all 1,100 local units, including counties, cities, schools, and libraries.Currently, designing and building two mission critical applications to assess utility and rail property, a multi-year project worth hundreds of millions of dollars to the State and taxpayers. This multi-year project replaces the updates I made to these applications over the years and is another step our agency has taken to streamline the property tax system for both personal and business taxpayers while maintaining the highest level of transparency.

Indianapolis, Indiana, United States

Unfortunately, I experienced such harassment at this role that I quit and consulted a lawyer. I am sharing this information to explain the quick turnaround and do not wish to speak about it further.

Texas, United States

This was a part-time role I took due to the difficulty of finding full-time employment during the pandemic. It was tremendously rewarding, and a tremendous challenge.Taught Python for data science remotely to students at Southern Methodist University in Texas. Topics include numpy, pandas, matplotlib, API design with Flask, and both SQL/NoSQL databases with sqlalchemy. Also teaching online dashboard design with Javascript, including Bootstrap, Plotly, D3, Leaflet, and GeoJSON; Tableau; and an introduction to machine learning in Python with sklearn.I'm proud to say that 7 of my 19 students have received either promotions to analytics positions or have found full-time work in data analytics or data science.

Indianapolis, Indiana, United States

NOTE: I was working in this role when the pandemic began. They ended my contract early in order to concentrate on other matters.Designed and executed Talend jobs to transform data to meet OMOP Common Data Model standards for healthcare research. This included reviewing packages written in PL/SQL and T-SQL and then translating those routines into a Talend workflow.

Carmel, Indiana

Administered IBM Private Cloud for Data (ICP4D) Environment. This included integration of MISO resources with LDAP accounts and creation and administration of MongoDB, Ceph, and other object stores on both virtualized and hybrid cloud architectures.Designed Flask API to communicate with MySQL,Netezza, and other MISO data environments. This included developing the API structure in OpenAPI and programming the service calls using Python.Note that this was a three-month contract with a one-month extension, with no renewal option. I took it as a stop-gap role when the CEO and CFO of Cloudera publically declared their lack of faith in their Hadoop offering, which ended my MPH contract and that agencies' Hadoop platform in general.

Indianapolis, Indiana Area

Administer 21-node production Hadoop cluster. This includes installing Anaconda, Python, and PySpark on all nodes; facilitating the development of machine learning jobs using Hive and Spark 2; installing JupyterHub; implementing the Department of Defense’s OSCAP security standard on all nodes; and designing an end-to-end security framework utilizing TLS, Kerberos, SELinux, and Sentry role-based security. As a result, the time needed to link records from all partner agencies was reduced from several hours to 13 minutes.Administer 7-node development Hadoop cluster. This included upgrading Cloudera Manager and CDH; configuring TLS, Kerberos, and data-at-rest-encryption; installing Spark 2 and HBase; and installing and configuring Sentry and Cloudera Navigator.Administer RStudio and JupyterHub on multiple servers, allowing data scientists to use multiple R and Python versions.

Indianapolis, Indiana Area

Developed supervised learning model to categorize retailer products using Python. This model was designed to replace services provided by an external vendor and had an accuracy score of 93%.Developed custom MS Excel Add-In to validate retailer product URLs. This was used by data analysts and replaced a manual “cut and paste” solution.Developed custom workflow using SQLite and PowerShell scripting to onboard new retailers and apply standard validity checks while preserving audit trail and allowing for batch processing.Administered SAP Hybris e-commerce platform, including adding new retailers, loading and reconciling retailer inventories, and running and troubleshooting scheduled jobs.

Indianapolis, Indiana Area

Developed custom solutions for clients using Hadoop, Hive, HBase, Impala, Cloudera Manager, and Talend's Big Data Platform, Data Quality, and Data Stewardship tools. Currently working under an NDA with these technologies and traditional RDBMS such as Oracle and SQL Server.Mentored future data scientists for the Thinkful coding boot camp. Instructed students in using Python for data science, including Numpy, Pandas, and SQLite. Taught statistics for data science, including sampling, normality, tests of significance, regression, and machine learning algorithms such as Bayesian classification, random forests, principal components analysis, k-means clustering, and neural networks.

Indianapolis, Indiana Area

Oversaw entire lifecycle of all Department data, including student ISTEP test scores and data used to evaluate teachers and school corporations. Directed data ingest, storage, ETL, analysis, and release workflows. Negotiated contracts with testing vendors. Coordinated public release of high-interest data such as student test scores. Ensured compliance with federal regulations (FERPA), statutory deadlines, and security best practices (NIST).Supervised build, configuration, and deployment of 21-node Hadoop cluster to centralize data from Oracle, SQL Server, proprietary files, and other data stores. Directed full-time and contract team members across multiple agencies and negotiated contracts for server hardware, Cloudera Enterprise, and the Talend product suite. Held weekly SCRUMs and coordinated training of existing development team in Big Data fundamentals, Hadoop, Hive, Sqoop, Hbase, Talend, Zeppelin, and Linux. Mentored team members and resolved technical and network issues to ensure continued productivity and quality work.Developed Hive-based data warehouse with customizable table schemas to meet data needs of diverse clients. Supervised development of Sqoop framework to update warehouse data from SQL databases incrementally and on-demand. Created standardized HBase “product sets” to aggregate complete data domains (e.g. student, teacher, school) into comprehensive profiles accessible through a common presentation layer.Developed workflows to ingest data from Oracle, SQL Server, Excel, numerous proprietary formats used by testing vendors, and other data stores into Hadoop data lake and Hive warehouse. Designed model ETL, ingest, analysis, and reporting jobs in Talend Big Data, Data Quality, and Data Integration, and supervised their implementation using standard and custom Talend components, Java modules, joblets, and Linux shell scripts. Managed repositories of re-usable Talend contexts and schemas for development, staging, and production environments.

Indianapolis

Supervised analysis of assessment data from 92 Indiana counties to ensure compliance with state property tax standards. Directed annual reviews of county sales data, utility property, and railcar property. Met with county assessors, tax lawyers, and legislators to develop assessment standards and negotiate tax liability.Designed valuation metrics and developed programs in R, SPSS, and Visual Basic to ensure efficient calculation. Programmed database to match property assessments with relevant sales data. Built data warehouse of construction labor rates using optical-character recognition (OCR) of public works contracts. Wrote SPSS scripts to automate routine statistical analysis of assessment data.Developed valuation models of residential, commercial, and agricultural property. Calculated cost schedules using cubic splines and nonlinear regression. Predicted data arrivals using non-stationary Poisson process. Adjusted hypothesis tests using modified Bonferonni procedure to control experiment-wise error rate.

Master Of Arts (Ma), Political Science

Degree earned in 2005 in PhD program (unfinished).

Bachelor Of Arts, Political Science