Wesley Turner Email & Phone Number

Seattle, Wa, Us

Houston, Texas, Us

Platform for managing deployments of ML models atop Kubernetes.

Houston, Texas, Us

Tools, improvements, and integrations on the client side for our model-training platform.

Mountain View, Ca, Us

I developed proof-of-concept solutions for clients' business problems using machine learning and the Google Cloud Platform.Clients who represented their business but who were not themselves machine-learning experts would come with a general idea of an opportunity. At a project’s start, I’d help them frame their business needs as a problem that can be addressed with AI. We’d discuss input data and methods to evaluate the relative value or success of any solution.My team and I then put together reports describing the problem in as technically-detailed a fashion as necessary to de-risk the project, to drive agreement amongst all parties, and propose a general approach to solving that problem. I’d stay in contact with our client stakeholders, sharing our progress, suggesting strategies for resolving the emergent ambiguities, training them on the software we're developing, and teaching them about how to think about ML just in general.On the more technical side, I led teams of sub-contractors to implement our proposal. Depending on the job and the size of the team, I might write nearly all of the code for the project myself or just provide the project's code review while drafting and reviewing technical designs.

Mountain View, Ca, Us

I began this job adding features to and generalizing our team’s simulator of the Google supply chain. But when I realized that we had neither the ability to measure our simulator’s fidelity nor evaluate the effects of policy changes we’d make to our build pipelines, I pivoted to a role centered around gathering data and exposing metrics myself, and helping other teams within Supply Chain do the same. I rectified data across different sources, defined metrics, and built a dashboard tracking statistics that were followed in our org’s first QBR.

Us

Skunkworks for senior leadership in the Oracle Data Cloud. I began this job evaluating the feasibility of one of leadership’s pet projects. After building a prototype model and sharing a report, I stayed on in the role, finding other areas within the organization in which data could tell a story that affected people's understanding of the business. I’d meet with interested parties, write software to gather and analyze data, and produce conversations and written reports with recommendations for directions the business should go.

Us

I developed analytics products to give advertisers insights into how their campaigns perform with different kinds of customers, and that surface new, potentially rewarding market segments for them to target. I focused particularly on what our data meant to customers and what they're able to do with it.I also helped incrementally change the culture away from one where engineers owned just the implementation of software they were directed to make and toward one in which engineers owned the entire product. I encouraged individual engineers to spend time thinking about how their products helped customers, and to suggest their own improvements. As a result, we built an "innovations" board that both gave ICs a greater feeling of ownership and won time for self-directed projects.

Seattle, Wa, Us

I designed, built, and deployed a framework to test changes to “Customers who bought also bought” pairings that allowed my team to detect even a 1% change to their efficacy within a single day – an order of magnitude more power than we would have had using the Amazon-standard testing framework. Before my software, our team could only test one or two changes per month, and so relied on intuition to drive most of our changes. This change gave us the ability to run several tests per week, and so drove a cultural shift within our team to use statistics when deploying changes to our similarity metrics and algorithms.I also:* Spread the experimentation technology to several other teams around the company, including to the teams making personalized recommendations and merchandising campaigns on the gateway.* Redesigned and implemented the build and service for the system that allowed customers to browse vendor-approved accessories for their items.

Seattle, Wa, Us

Experimented with the addition of new types of customer behavior data into search results.* Prototyped the first hybrid catalog/behavior-driven search engine. Before this experiment, customer behavior had only been a process run on top of and separate from classical search.* Implemented and deployed new categories for top-seller-like rankings. New categories were "Most wished for", "Most gifted", and "Most highly rated".

Phd Candidate, Applied Math

Mse, Computer Science

Bs, Math