David Dodds Email & Phone Number

Mclean, Virginia, Us

• Consulted federal clients with mortgage data engineering [US CFPB] and cybersecurity data engineering [US CISA]• Reconciled entities in the Panel of Institutions dataset, for the CFPB and partner agencies [Federal Reserve, FFIEC] • Published financial metric dashboards via Quicksight and serverless Athena queries, saved business analysts’ time• Mapped changes to cybersecurity metrics, co-authored proposal to deploy Amazon API Gateway for vendor dataKey Achievements► Oversaw updates and reconciliation of financial institution identity datasets, moving the reconciliation platform to a higher level of automation than previous years. Each data point reconciled and updated is estimated to be worth US$15k-$350k in downstream economic value, depending on whether it is used by regulators or longitudinal market researchers. Thousands of such data points were reconciled in a semi-automated fashion yearly.► Worked with a team of data engineers and business analysts to move filing reports of all mortgage applications in the U.S. (12-26 million records annually) away from Excel and into serverless-query dashboarding, an intermediate level of automation towards internal user-interactive dashboards. ► Deconstructed the financial institution identity reconciliation process to various levels of granularity, providing diagrams to support its ongoing redesign and automation. Several co-workers in the department claimed they finally understood it after years, because of my diagramming. This documentation is still referenced as they continue to automate the entity reconciliation codebase. ► Organized and simplified client’s data team work tracking boards, with an emphasis on time scope, workstreams, filtering, and the relationship between developmental and operational projects. This prevented wastefully redundant work delegation, eased task juggling between emergencies and proactive builds, and created space for future-oriented projects to develop.

Tulsa, Oklahoma, Us

• CA used postback data from partnered brands to orchestrate lead gen revenue and consumer matching analysis• Postback depended on data pipelines to ingest, transform, and organize external data into Redshift and MDM• Maintained the postback data pipeline DAG's in Apache Airflow via Python, SQL, and code in shared repositories• Freed data labour for warehouse migration to Snowflake, attended Snowflake workshops, assisted migration tasks• Maintenance of postback pipelines ensured and verified the flow of multi-million-dollar revenue monthlyKey Achievements► Maintained the data pipelines that ingested external partner sales data, while the data team was undergoing a data warehouse migration and development of company data governance. This data was directly used to bill partner companies and sell the value of CA’s brand recommendations funnel to other consumer-vetted brands.

Washington, D.C., Us

• TRP needed neural networks to collect recyclable items data at scale, from non-standard fragmented municipal sources• Used custom OCR software to label the data neural networks needed to automate recyclable materials data generation• Generated categorical recyclable materials data from thousands of municipal guideline sites from around the nation• Resulting neural networks gave TRP a reliable future data stream of frequently changing recycling market changes

Denver, Co, Us

• Audited non-standard fields that disrupted tracking of hosting reservations in Compass Mining’s NoSQL database• Cloned staging collections to test standardisation scripts, summarized collection structure with dictionaries, pandas• Advanced Compass Mining’s progress on an impending migration to relational and GraphQL data storage

Miami, Fl, Us

This role was via the Civic Innovation Corps for the City of Miami’s Department of Innovation & Tech [“DoIT”]. DoIT leads Miami’s data driven decision-making and open tech initiatives.• Created models for zoning inspectors, business owners to predict licensing times for 2,000+ new businesses yearly• Iterated tree ensembles, timeframe selection, to yield models outperforming human estimation 5x• Deployed a tool that prevented business failure, unplanned expenses, lost revenue for entrepreneurs and MiamiKey Achievements► Produced statistical learning models that provided more accurate business licensing wait time estimates to home-grown business owners and entrepreneurs, improved by a factor of 5 over human baseline estimates. Improved estimates saved local businesses tens to hundreds of thousands of dollars in lost operation time, rental signing, and inspection fines.

San Francisco, California, Us

AssemblyAI provides the most trusted deep learning transcription API for developers and startups in production-AssemblyAI needed how-to Python blogs showcasing their API, so prospective client developers could explore it firsthand► Distilled AssemblyAI’s documentation into actionable instruction, producing their debut article for connection with Python- Troubleshot novel connections with AssemblyAI to Zoom, YouTube API’s, delivering succinct instructional content for both- Enabled developers to experiment with world class speech-to-text software, amplifying AssemblyAI’s subscription pipeline

Washington, District Of Columbia, Us

• Contributed to a team that uses open data to disrupt illicit trade networks and global security threats• C4ADS created SanctionsExplorer, an open international sanctions search engine of U.S. Treasury, U.N., E.U. data• Measured entity fragmentation, fraudulent aliases (persons, organisations, vessels, aircraft) in SanctionEx database• Prevented SanctionEx users from making illegal transactions or losing business opportunitiesKey Achievements► Presented use cases and tradeoffs to guide C4ADS’s decision whether to adopt a COTS API service to overcome internal data silos. ► Produced first accuracy report of a database used globally by financial institutions, nonprofits, and transnational corporations to avoid extremely costly business deals with internationally sanctioned entities.

This action-based Software Entrepreneurship course develops leaders who create software that solve costly problems in niche industries.

Supported teachers and pre-kindergarten children in pro-social, project-based emergent learning. My goal is to reform education systems, therefore I must experience excellent educating on the ground, where the rubber meets the road, in a living breathing school community!

Data Science

Bachelor Of Arts - Ba, Elementary Education And Teaching

Piscine