Staff Machine Learning Engineer, Traffic Intelligence

<div class="content-intro"><p><span style="font-family: helvetica, arial, sans-serif; font-size: 12pt;">Airbnb was born in 2007 when two hosts welcomed three guests to their San Francisco home, and has since grown to over 5 million hosts who have welcomed over 2 billion guest arrivals in almost every country across the globe. Every day, hosts offer unique stays and experiences that make it possible for guests to connect with communities in a more authentic way.</span></p></div><p><span style="font-family: arial, helvetica, sans-serif; color: #000000;"><strong>The Community You Will Join: </strong></span></p> <p>Our web and API surfaces handle requests from guests and hosts alongside a growing volume of automated agents: AI assistants, crawlers, and scrapers. We build the systems that bring clarity to this traffic, combining in-house ML and vendor signals to decide in real time how to serve billions of daily requests. Anti-bot and anti-scraping detection is our most adversarial mandate, but the wider challenge is full traffic classification: building evaluation frameworks that tell legitimate automation apart from abusive actors, so high-stakes decisions hold up across the fleet.</p> <p><span style="font-family: arial, helvetica, sans-serif; color: #000000;"><strong>The Difference You Will Make:</strong></span></p> <p><span style="color: #000000; font-family: arial, helvetica, sans-serif;">You will architect and maintain Airbnb’s end-to-end traffic classification ML systems, balancing high-performance model deployment with rigorous offline data pipelines. Success is measured by your ability to harden edge-traffic policies—targeting reduced bot-incident MTTM—and by establishing rigorous evaluation practices that ensure foundational signal accuracy and evasion-resistance across the fleet.</span></p> <p><span style="font-family: arial, helvetica, sans-serif; color: #000000;"><strong>A Typical Day:&nbsp;</strong></span></p> <ul> <li>Own the complete lifecycle of traffic-scoring models, from problem framing to real-time deployment, managing the adversarial feedback loop to ensure high evasion-resistance and directly drive reductions in bot-incident MTTM.</li> <li>Architect robust offline-to-online pipelines that produce certified source-of-truth datasets, establishing rigorous evaluation frameworks—such as stratified benchmarks and leakage-prevention checks—to ensure every model improvement is empirically measurable and defensible.</li> <li>Execute model optimization within strict millisecond latency budgets at the internet edge, uniquely balancing inference costs against incremental value while maintaining fleet-wide fail-open behaviors.</li> <li>Partner daily with