John Walk Email & Phone Number

Stockholm, Stockholm County, Se

ML engineer supporting search, content discovery, and personalization efforts in the Pluralsight user experience on a cross-functional team including data science & machine learning, full-stack engineering, and UI/UX design.From July 2021, also served as technical lead for the Discovery team's data streaming efforts.• Rearchitected Pluralsight’s search backend to better integrate the core Elasticsearch document store and ML-powered experiences with internal and external consumers while maintaining resilience under heavy load• Oversaw productionalization of graph-powered experiences driving topic discovery and exploration alongside the core Search experience• Overhauled ML model implementations in search-result reranking and content browse experiences for better stability, performance, and maintainability• Introduced scalable, distributed data processing with Dask to the company’s data ecosytem, powering some of the largest data pipelines in the company’s usage and driving user-facing experiences• Prototyped improvements to data processing architectures to streamline ingestion of disparate content types into our search engine• Developed reusable project templates for Python packages and FastAPI webapps, shared company-wide

Boston, Ma, Us

Technical team lead on the Merchandising Ops Data Science team, providing data-driven solutions for managing information on millions of Wayfair.com products.- Led technical efforts on the Tagging team, developing ML models for product categories & individual attributes, driving automated annotation for new products as well as correction in the existing catalog based on rich text and image information- Developed product roadmaps for scaling new tagging efforts to rapidly expand coverage of individual attributes, as well as for cross-class models touching on error flagging, schema design, and customer feedback data- Served as an engineering SME on the data science team, including code & design review, regular office hours, and the development & rollout of team-wide trainings to over a hundred Wayfair data scientists. As part of the training, developed an automated scorecard for development environment setup for new hires.

San Francisco, Ca, Us

In-depth alumni advising for the Spring 2018 cohort of Insight Data Science Fellows during their interview phase

Boston, Ma, Us

Senior data scientist on the new DS team at Cinch, building the company's capabilities for data-driven decisions and products. Cinch is a one-of-a-kind AI financial advisor, aiming to bring unbiased advice to users' personal finances to level the playing field with banks and credit card issuers. - Developed ML model pipelines supporting financial strategy recommendations, covering the full stack from raw data ingestion to production implementation - Implemented "data lake" storage, processing pipelines, and shared modeling resources/environment for DS team to streamline data analytics and model development - Designed numerous analytics dashboards & visualizations for BI and data analytics using open-source tooling, both using self-contained solutions (Metabase, Superset) and a custom Ember.js webapp featuring Bokeh visualizations for more fine-grained control over the design - Compiled custom training materials in SQL, Python, statistics, and other data tools for junior business/data analysts

San Francisco, Ca, Us

Insight is a brief, intensive postdoctoral fellowship in data science, bridging the gap between STEM academia and industry. Each fellow builds and presents a tech-demo data science solution for real-world problems, developed end-to-end over the course of the program. The Boston session particularly focuses on healthcare & biomedical applications. - Designed a sentiment analyzer & recommender engine for antidepressant usage and preferences based on social media discussions on reddit.com. - Scraped input data from ~29,000 relevant reddit user comments with reddit API, parsed into SQL database. - Implemented TF-IDF scoring-driven conversation characterization for each medication, capable of finding highly-specific points (for example, finding off-label usage as a smoking-cessation aid for one drug) - Developed a sentiment analyzer utilizing parsed tree structure of sentences to intelligently "chunk" comments around mentions of single medications, account for semantic information in the sentiment score to gauge user satisfaction with medication(s) - System recommends likely successful alternate medications presented through webapp interface

Cambridge, Ma, Us

Experimental and computational researcher on the Alcator C-Mod tokamak, a major DOE project for fusion energy research, in the pedestal physics group.- experimental and computational modeling to understand the plasma edge in high-performance scenarios, both conventional concepts and the novel "I-mode" regime pioneered at Alcator C-Mod- development team for new modular, extensible data analysis toolkits in Python for fusion research

Cambridge, Ma, Us

Experimentalist on the Alcator C-Mod tokamak, a major DOE project for fusion energy research, in the pedestal physics group.- maintained and operated core and edge Thomson Scattering spectrometer systems during experimental campaigns- research focused on edge structure and stability in high-performance reactor regimes- research presented in a number of academic conferences and peer-reviewed journals

Cambridge, Ma, Us

data analysis and operations for X-ray spectrometer arrays on the Alcator C-Mod tokamak.

Doctor Of Science (Sc.D.), Nuclear Science & Engineering

Bachelor Of Science (B.S.), Physics And Mathematics