Stuart Owen Axelbrooke

Senior Machine Learning Engineer

(206)334-5138
stuart.axelbrooke.com
Github: soaxelbrooke
LinkedIn: stuartaxelbrooke

Past

Senior Machine Learning Engineer with 12 years experience including Amazon and Startups, applying Machine Learning, Cloud, Kubernetes, and Big Data technologies to deliver > $100M in annual revenue increases and significant boosts to user retention.

Future

Seeking Senior/Staff Machine Learning Engineer positions leading tech, growing/mentoring junior and mid-level engineers, and delivering high-impact solutions and architectures in ambiguous problem spaces.

Experience

Amazon Senior Machine Learning Engineer (L6)
2020-07 to 2022-12 (2 years, 6 months)

Lead Relevance and Trust/Qualification pillars for Amazon's Choice (AC), improving the quality of recommendations.

  • Proposed, designed, trained, and productionized a tabular and text-based relevance classifier
  • Increased annual revenue by $180M
  • Reduced returns on AC purchases by 3.5%
  • Reduce AC relevance defect rate by 25%
  • Authored tech roadmaps and architectures, reducing system complexity and improving org velocity
  • Mentored junior engineers and created/ran tech talk series for professional growth and knowledge sharing
  • Lead team of 8 engineers delivering ML, pipelines, and API projects
Thought Vector Founder
2018-01 to Present

Solo-founder at Thought Vector, building out ML products and providing ML consulting.

  • Builds Taggit, a no-code text classification platform
  • Built automatic model training, model eval, active learning, and online inference systems
  • Built full-service text-tagging and dataset management system
  • Built backend (Python/PyTorch) and frontend
  • Taggit targets feedback quantification, content moderation, topic modeling, etc applications.
  • Delivered object detection/image classification system for identifying specific shoes from consumer photos
  • Built smart config-driven scraper for continuous and efficient data collection from Apple, Twitter, etc
Apptentive Senior Machine Learning Engineer
2015-06 to 2018-01 (2 years, 7 months)

Spear-headed creation of data science and machine learning products.

  • Delivered NLP-powered Insights product providing text exploration, boosting dashboard MAU 150%
  • Designed and delivered Love Report data product, forming basis for white glove data services
  • Delivered quick-response predictor adopted by largest customers to reduce support costs
  • Proposed and delivered phrase extraction/topic model
  • Proposed and delivered sentinement classifier
  • Coordinated with product and customer success teams to define experiments and features
  • Lead teams of 2-3 engineers to deliver data/ML projects
Informion Founding Machine Learning Engineer
2014-12 to 2015-06 (7 months)

Built out data pipelines for product recommendation push re-engagement SaaS product.

  • Built candidate set via information retrieval and full-text search techniques
  • Created search-based personalization/recommender system
  • Created engineering process reducing incident rates significantly
  • Implemented recommendation quality metrics
  • Lead team of 2 engineers
Emotiv Labs Co-Founder, CTO
2014-03 to 2014-12 (10 months)

Built group gifting platform GiftStarter, including SPA front-end and real-time slot buying scalable backend. Managed 1 other backend engineer.

  • Built real-time slot reservation/selling system on Google Cloud Platform
  • Authored Javascript front-end and initial Python back-end
  • Created embedded tag for e-commerce partners to enable cross-site group gifting
  • Lead 1 back-end engineer
TRUE STYLE Founder
2013-08 to 2014-03 (8 months)

TRUE STYLE was a personalized men's fashion shopping experience that learned user preferences.

  • Built product and brand recommender system based on user behavior
  • Integrated with external affiliate marketing systems
Cypress Semiconductor Senior Applications Engineer
2010-08 to 2013-08 (3 years)

Designed and implemented complex signal processing systems via Cypress' programmable system-on-a-chip for large scale external customers. Applications include cardiac-wall conductance measurement, low-power sonar, and 3D-touch mobile touchscreen measurement.

Papers and Patents

Patent Offloading tasks from a central processing unit to peripheral function engines: A method for maximizing utilization of on-chip SOC resources by offloading tasks to heterogeneous co-processors. Reduces overall BOM costs and system power consumption.

Projects

NLSea + AToM - Seattle ML/NLP Meetups Organizer

python-bpe - Byte Pair Encoding in pure Python

phrase - Super fast phrase extraction tool

Customer Support on Twitter Dataset - Conversational dataset of high value business-to-customer interactions

Skills

Education

BS Electrical Engineering from Washington State University, graduated 2010.

Fun Facts