Arthur Caillau

Early Trust Engineer at Airbnb. I work across teams, products and functions: from Software Engineering, Data Engineering to Machine Learning Engineering. I value high quality code and advocate functional programming techniques. • Designed, implemented and productionized online NLP Deep Learning models at scale to detect offline incidents using Tensorflow, Keras and internal services • Helped reduce Account Take Over (ATO) rates by 85% by fixing fraud loopholes, launching 2FA on untrusted devices and deploying ML models • Fixed, improved and implemented critical data pipelines that power up online ML models, metrics and dashboards • Revamped email and SMS account alerts • Audited Risk related codebases to find fraud loopholes and improve observability • Launched internal web tools for TnS agents to investigate and mitigate fraud cases which leads to a 20% average handle time improvement • Designed and implemented full UI session tracking for internal web tools

Early Software Engineer at Stuph. • Implemented the webapp - backend and frontend - using Clojure/ClojureScript - including designing API endpoints, web scraping and email scanning • Built an automated email scheduler and tracker • Designed data extractors, parsers and scrapers to collect and organize data on users. • As part of the founding team, worked extensively on figuring out the product, and strategies for growth and marketing

Designed a web document management system.