Josh Dubose Email & Phone Number

Braunschweig, Niedersachsen, De

Collaborate with business stakeholders, data science, and data engineering teams to develop and deploy operational machine learning products across key functional areas, including vehicle pricing, credit decisioning, fraud detection, vehicle recovery, and sales forecasting- Partner with business teams to identify challenges, diagnose problems, and build advanced analytic products that enable business transformation- Lead executive-level presentations, simplifying complex data science projects into business capabilities- Leverage foundational knowledge of machine learning principles and algorithms to inform model design and deliver robust, scalable solutions- Functions as a Data Scientist to develop and implement solutions in the Operations Research space such as forecasting, simulation, and optimization

London, Gb

Developed the Federal Audit Analytics strategy across 4 clients and drove cross-project collaboration and efficiencies on data processing, preparation, enrichment, storage, automation, and visualization initiatives- Managed the design and build of a transactional database in Azure Cloud to empower audit teams with centralized access to transactional level data and Business Intelligence reports - Functioned as the analytics Subject Matter Expert during technical code reviews to augment back-end approaches, polish scripts and stored procedures, and work through data complexities and challengesLed the assessment, design, development, and implementation of a production analytics environment and dashboard capability for the Department of Commerce headquarters and finance function- Coordinated the extraction of over 200 source tables from multiple legacy systems to create a single analytics sandbox within an on-premise Microsoft SQL Server database- Enabled cross-system and business function analysis that was not previously possible and uncovered operational insights that led to data-driven internal controls recommendations and the ability to monitor internal control risks

London, Gb

Supported Internal Controls by assessing a legacy Oracle database to understand system design, explore data availability, uncover and identify system and process gaps, and to unify segmented data for aggregate reports- Developed SAS scripts to perform data quality enrichment of free text data fields that are critical to asset valuation by leveraging alternate data sources, developing business rules, and utilizing matching algorithms – resulting in an increase in known asset numbers from 28% to 98%

Reston, Va, Us

Managed development of a custom java application with risk scoring functionality that informed port surveillance of high-risk products being imported- Drove transformation of the risk scoring methodology from a business rules design to a logistic regression modeling approach to increase accuracy of predicting product non-compliance- Managed the development team to ensure tasks were completed on schedule and that functionality met design- Engaged with end-users to elicit feedback and develop technical requirements for enhancements to application

Armonk, New York, Ny, Us

Led data mining effort to identify mailer short-payments directly enabling USPS to recover $20 million in 2013 - Utilized SAS to clean, manipulate, merge, analyze, and aggregate data incorporating the use of do-loops, arrays, and macros to facilitate analysis of mailer manifest data and to reconcile data that ‘errored-out’ of the framework of the IT procurement system - Established business policies and defined business requirements to mitigate payment gaps within the procurement system and represented USPS in presenting these changes to the Mailer’s Technical Advisory Committee- Led response to Office of Inspector General, through data analysis and a deep understanding of the Mail Entry and Payment business, by building a case defending USPS against a report of revenue gapsDeveloped Discrete-Event Simulation models to optimize resource allocation and to predict operational performance for USCIS- Designed and managed model process maps - coordinated with multiple client working groups and internal solution architects to align model representation of a complex IT solution and client business processes- Developed the simulation model and performed model verification and validation for numerous ‘what-if’ scenarios- Implemented an optimization algorithm to automate a previously manual process of model resource allocation- Authored and presented Project Findings Report – primary project deliverable including project scope, approach, methodology, analysis, projected cycle times, and optimal resource allocation

Wildlight, Florida, Us

- Created Process Safety Management (PSM) inspection routes specific to OSHA hazardous chemicals and Preventative Maintenance (PM) routes to establish statistical norms and set equipment life-cycle alarms- Built, managed, and updated PSM and PM databases- Integrated routes with industrial PDA uploaded to PSM and PM databases, enabling documented safety measures and equipment life-cycle data

Detroit, Michigan, Us

Designed and implemented a standardized, departmental shut-down process to reduce electricity usage during plant downtime