I served as Head of Technical Product Management, Enterprise Data Governance Technologies, and Adoption & Enablement for the Unified Data, Analytics, and AI/ML Platform (UDAIP) which represents a comprehensive, holistic solution for the collection, management, and use of data at US Bank. I led an organization of 30+ dedicated individuals which I built from the ground up, and received two major promotions in my 4+ years at the Bank.Data flows into the UDAIP from many sources ranging from traditional Systems of Record to customer actions / events via Digital Experiences to external data sources. Depending on the nature of the data / events, updates may come in via either the Batch Data Processing capability (metadata-driven platform developed using Spark and Airflow) or the Cloud Digital Artery, a low-code / no-code streaming platform built on Confluent Kafka.In addition to a modern online serving layer using Cassandra and GraphQL API’s, the platform includes a Workbench / Data Catalog to discover and govern data, a cloud-native Data Lake / Data Warehouse platform, and provides self-service BI / Analytics, as well as AI/ML model development and execution.The UDAIP is built to be cloud-centric from the ground up, with some components being built directly on Microsoft Azure, while others were built on-prem or on GCP and have been migrated or are in the process of migrating to Azure.As just one facet of the impact provided by the platform, the Online Serving Layer alone is already generating annual savings in the $15-20MM range of cost avoidance from mainframe MSU’s, as well as powering many new / growth use cases (such as allowing a ten-fold increase in velocity of marketing campaigns, powering $34MM in incremental revenue).Another example of my team's impact is both building and administering a number of the technologies and processes supporting the controls required to enable the Bank to migrate real customer data to the cloud.

Following the acquisition of Yahoo by Verizon, my role expanded into full Product Management responsibilities for URS (the User Record Store), which comprises the repository of all data within Verizon Media for user understanding and the tools to make it discoverable, understandable, and agile. URS is truly "open-source data", meaning I interact with customers both upstream and downstream in every corner of the company, from Ads and Targeting, to Media and Communications products, to Customer Care and Abuse, and so on. Products I own under the URS umbrella contribute in the neighborhood of a billion dollars of revenue under influence for Verizon Media, help power personalization that increases click-through rates on articles by 20-30%, facilitate abuse / criminal investigations, help provide differential white-glove treatment for our most valuable users, etc.In addition to sustaining and growing the impact of URS, I am also building an entirely new product LUUP (Longitudinal Unified User Profile) from idea to reality. LUUP compliments URS by lengthening the time dimension of our user understanding, and blends signal from multiple sources to give a "single best answer" about many facets of our users (everything from what sports teams, public figures, etc they want to follow to what product categories and brands they are interested in and what chains and categories of stores they tend to frequent, etc.). Constructing these user attributes requires heavy use of machine learning techniques and I have four different science teams within Verizon Media that I am partnering with to build this product.

Straddling the technical / functional divide, I assumed responsibility for defining and annotating schemas to provide structure and both human and programmatic discoverability to large volumes of data from disparate sources centered around the theme of user understanding. The User Record Store (URS) was established to ingest and enrich, as well as make discoverable and understandable user data (user actions, declared interests, demographic data, etc.) from essentially every Yahoo product at the time (News, Sports, Finance, Tumblr, etc) as well as Display, Native, and Search ads. Historically these were in varying locations, with no consistent or well-documented structure, making it difficult for data to be leveraged across teams and nearly impossible to bring data from different sources together to gain deeper insights. Included in my role is to bring rigor and as much standardization as possible to how we structure the data and evangelize that schema should be treated with the same level of respect as code, as well as to curate metadata with sufficient quality and depth to enable potential consumers to not only find data of interest for their use cases but moreover to unlock previously unobtainable value by joining together data from different sources.As the URS and Geo teams were sister teams at this time I also played a data architecture and product management role for Yahoo’s Geo team, and in synergies between URS and Geo. For example, I brought to market the Best Known Geo product which established the canonical source of users’ Home, Work, and other relevant locations. I also spearheaded the leveraging of users’ actual presence at venues such as sports stadiums to serve as a confirmatory signal to users’ interests in Sports teams.

I led a turnaround effort for a categorization system that is a critical part of Yahoo!'s ad targeting ecosystem, a database which had a vast array of tables, jobs, even entire schemas of unknown history and which no one could say for certain were even used or not, and which served as backend to a UI which in some areas was so non-performant as to be literally unusable, with some screens consistently timing out before they could even be rendered. Through research and reaching out to a wide array of external teams, I was able to shut down or drop many of the extraneous schemas/objects/jobs and my extensive performance tuning efforts led to such a dramatic change in usability of the UI that it was termed as downright "zippy".Next I turned my attention to some highly problematic data architecture issues such as circular replication flows between Targeting's two Oracle databases as well as use of one of them as a pure pass-through recipient of materialized view data which another database queried via dblink. The two databases were consolidated in a phased approach--first physically, then logically--in order to split the risk and disruption given the challenges both technical and organizational.Over time I further evolved into a data architect role helping fill a void with respect to understanding of end-to-end data flows and even delving into areas such as problem management when I see a gap/exposure. I owned the Targeting piece/impact of several wide-ranging initiatives such as the sunsetting of Yahoo!'s legacy booking system and rearchitecting of APT to leverage data published in object-based rather than pure relational format, and the database portion of our integration with Interclick which includes establishing heterogeneous replication between MSSQL and Oracle. These projects were in addition to continuing to provide ongoing support such as design and code reviews for the database component of all development work by nine scrum teams across three continents.

Assumed full ownership of database engineering work for a complex migration project for advertiser accounts which represented a critical success factor of the Y!-Microsoft Search Alliance and which required integrating nearly a dozen systems and similar number of teams in multiple geographies. From there, evolved into a lead role with responsibility for a portfolio of several systems including Ad Targeting, Inventory Management, and Opportunity Management/Sales Tools. For each of these properties, I was responsible for ensuring scalability, supportability, adherence to best practices, etc. Specifically this entailed full ownership of all database schema objects and changes managed via a robust build system, as well as performance tuning, consultative/educational efforts on Oracle/RDBMS technology and concepts, and facilitation of communication and translation of requirements between dev/product/qa teams and the operational/infrastructure-oriented DBA team. I also undertook numerous team-level/leadership roles such as defining an "onboarding" process for new properties to work with our team, developing a rubric for estimating level of effort for various tasks to aid in developing project plans, and helping establish/maintain an "architect-track" working group for mentoring and knowledge-sharing including both delivering tech talks myself and pulling in speakers/experts from outside groups within Y!.

Served as senior/lead DBA on a wide range of Yahoo! Audience and Marketplace properties, most of them high-profile, customer-facing systems. I was responsible for a stabilization effort for Y! Answers, which was deployed with speed-to-market in mind more so than scalability, but which almost overnight became the #2 reference site on the entire WWW behind only Wikipedia, leading to EVP-level visibility of even the smallest database hiccups. My contributions included extensive performance tuning; RAC, Linux, and NetApp architecture work; capacity planning; and many other technical and process changes. I also led a highly successful, from-the-ground-up deployment of two 10-node RAC clusters for a complete rearchitechting of Y! Personals (migrating from MySQL), taking over when the project was three months behind and the customer relationship with the DBA team damaged, and managed to finish over a month ahead of schedule and the customer delighted with the partnership. For each of these (and other) projects, I personally prepared, presented, and defended proposals for needed servers and filers before the Hardware Review Committee, which consists of one of Y!'s co-founders and other VP-level executives, with a 100% success rate. I also aggressively sought out opportunities to save energy and money by reclaiming dozens of unused or under-utilized servers and filers. Moreover I served as interface to Oracle to facilitate tracking and discussions of ongoing technical and support issues, and coordinated weekly executive status reports, redefining roles and responsibilities to streamline what had been a cumbersome process and resulting in higher quality, more complete reports.

Served as lead DBA for one of HP's most sensitive systems, providing single sign-on for many of the company's eCommerce assets, which employed technologies such as Real Application Clusters and Data Guard to meet its exceptional uptime requirements. After joining, I led a highly successful implementation of Oracle's Multi-master Advanced Replication to further enhance system availability. My efforts also enhanced overall robustness of the system through improvements in monitoring, security, etc.In addition to primary role as Oracle DBA, I also responded to organizational needs by serving as an MSSQL DBA for Microsoft's Sharepoint knowledge sharing product, as well as Change Manager, with responsibility for tracking, approving, and scheduling all proposed changes to the system as a whole (not just the database component).

As primary DBA for one of the world's largest PeopleSoft HRMS systems, I identified and solved numerous challenges, many involving performance tuning (just one example being a tax-filing extract, late delivery of which would incur a half-million dollar fine, which I tuned from 18 hours to 45 minutes). My contributions were recognized to the point I was elevated to a lead position, and then to acting manager for over half a year while the position was vacant. In this role, I oversaw a team of eight Oracle DBA's, set priorities, communicated regularly with management up to VP-level, etc.

Serving in a software development environment, in addition to operational DBA tasks, I had the opportunity to gain experience in areas ranging from data modeling (using tools such as Oracle Designer) to the "packaging" of releases (developing professional-looking scripts, documentation, etc. to roll out new releases to our client sites). On the latter, management noted that on the first release for which I was responsible for this packaging, we received the best feedback from our customers and fewest support calls of any release to date, despite it being the most technically complex we had yet undertaken.

My roles with Accenture ranged from hands-on DBA and other technical work, to technology strategy (e.g., evaluating products and asking the tough questions of vendors to design the overall technology and application footprint for a startup company), to project management/resource planning. I assumed my first team lead role after less than a year of total IT experience and was promoted from analyst/staff to consultant/senior in the shortest allowable timeframe. Gained invaluable experience and perspective from exposure to a wide variety of projects and industries and Accenture software lifecycle methodologies. Over the latter part of my tenure I specialized in Retek (now Oracle Retail) software and worked with several retail clients, including an instrumental role in implementing perpetual inventory for the first time at one of America's most admired and respected companies.