# Kevin Loftis > Software Engineer @ Whatnot Location: San Francisco Bay Area, United States Profile: https://flows.cv/kevinloftis Experienced software engineer with a machine learning/data science background helping companies build high-performance production ML systems. ## Work Experience ### Software Engineer, Async Marketplace @ Whatnot Jan 2025 – Present | San Francisco Bay Area In this role, I've owned 0→1 efforts for launching Whatnot's Short Video platform and Listing Comments system end-to-end — from product scoping and technical design through implementation, analytics instrumentation, and launch. For Short Video, drove 5x Short Video listing GMV growth and 3x video supply growth quarter-over-quarter (and accelerating). For Listing Comments, achieved 5x organic comment growth and 7x unique commenter growth month-over-month post-launch. ### Software Engineer -- Data Infrastructure @ Whatnot Jan 2023 – Jan 2025 | San Francisco Bay Area - Designed and built Whatnot's PlanetScale CDC pipeline, a real-time Change Data Capture system replicating production MySQL (Vitess) data to Kafka and Snowflake via Debezium and Kafka Connect. Authored the core RFC and technology evaluation, deployed the Strimzi-based Kafka Connect framework on Kubernetes with Helm and Argo CD, designed a declarative Data Contract configuration system embedded in schema definitions, and built connector deployment automation — enabling production tables to be replicated for analytics, search indexing, and business intelligence. - Engineered an Flink-based real-time event distributor application that routes analytics events to dedicated Kafka topics for downstream consumers, significantly reducing consumer load to only the events they need, eliminating event delivery delays, enhancing reliability through infrastructure-level isolation, and reducing Kafka network egress by 50%. - Re-acritected analytics events pipeline to shift event preprocessing upstream into a real-time Flink application reducing event ingestion delays into Snowflake by up to 20 minutes while also unifying event preprocessing across all event consumers. - Scaled realtime event pipelines to support 50x current throughput and reduced kafka infrastructure costs by 80% via implementing message compression and partition scaling. - Designed and implemented a partitioned, declarative training dataset creation framework enabling efficient, reusable training datasets—reducing compute overhead, eliminating data duplication via Snowflake external tables, and accelerating ML feature iteration. ### Senior Software Engineer, Machine Learning @ Reddit, Inc. Jan 2021 – Jan 2023 As a Founding Engineer of the Machine Learning Features Team at Reddit, I've made significant contributions to enhancing real-time ML systems. I led the creation of a high-performance streaming feature platform built on top of Flink, KSQL, Kafka, Redis, BigQuery, and Kubernetes, optimizing data availability for low-latency production inference and historical feature access. Additionally, my work encompassed the design and implementation of the Feature Store API, the development of a Python SDK for feature definition, and the creation of a feature metadata store to facilitate feature discovery and distributed access. ### Software Engineer II, Machine Learning @ Reddit, Inc. Jan 2021 – Jan 2021 | San Francisco Bay Area ### AI/ML Research Scientist @ Perceptronics Solutions, Inc Jan 2020 – Jan 2021 ### Data Scientist (Master's Practicum) @ Reddit, Inc. Jan 2019 – Jan 2020 | San Francisco Bay Area Streaming Feature Extraction Pipeline: Architected and developed a Flink streaming data processor in Scala. • Pushes comment and post related features to production feature store for use by models that control the popular feed. • Updates aggregate features in real-time Graph-based subreddit community detection: Developed a subreddit graph based on user view overlap and performed community detection on graph to cluster similar subreddits. • Implemented recommendation system that resulted in a 2x increase in subscriptions and a 15% increase in post and comment activity among subscribers; system developed in Python using subreddit graph-based features. • Conducted A/B test to lead product decision making around community recommendations. • Leveraged graph-based subreddit features to aid in manual validation of subreddit content tags reducing validation time by more than 97%. ### Senior Research Assistant @ Oregon Health and Science University Jan 2017 – Jan 2019 | Portland, Oregon Area • Generated segmentation of 3D image data for use in visualizations and training machine learning models. • Developed deep learning models for automated segmentation using Pytorch. • Led integration efforts of existing open-source software into research workflows allowing scaling of manual segmentation efforts. • Coordinated weekly machine learning journal club. • Produced animations and figures for publications. • Supervised interns ### Medical Informatics Intern, Department of Medical Informatics and Clinical Epidemiology @ OHSU | Oregon Health & Science University Jan 2016 – Jan 2016 • Led the redesign of clinical quality measure dashboard to improve data visualization and usability of dashboard interface. • Assessed improvement of dashboard design through user feedback interview and confirmed positive improvement. • Mapped patient-level EHR data to claims data and provided datasets to assess the relationship between risk prediction methodology and cost and utilization outcomes. • Conducted semi-structured interviews with clinic staff and assess the efficacy of the dashboard redesign. Analyzed qualitative findings indicating superior design and user experience as compared to the previous version. ## Education ### Master's degree in Data Science University of San Francisco ### Bachelor of Science (BS) in Molecular Biology Lipscomb University ### Post-Bac studies in Computer Science Portland State University ## Contact & Social - LinkedIn: https://linkedin.com/in/kevin-loftis --- Source: https://flows.cv/kevinloftis JSON Resume: https://flows.cv/kevinloftis/resume.json Last updated: 2026-04-11