Hire Xu D. - Forward Deployed Engineer

Serve as the primary technical advisor for Fortune 500 accounts on distributed data architecture strategy, working at the intersection of Trino query engine optimization and cloud-native storage (S3, Iceberg). Promoted twice. Architected hybrid data patterns and cost-optimization strategies that reduced customer compute waste across auto-scaling clusters Own severity-1 incident resolution for strategic accounts ($9M ARR), leading cross-functional war rooms with Engineering and Support Designed high-performance federated access layers enabling AI/ML teams to query raw operational data directly for context engineering — eliminating data movement bottlenecks Built and led a customer support engineering team, establishing escalation frameworks and mentoring junior engineers on distributed systems troubleshooting Was also Customer Support Team Lead and Sr. Solution Engineer before taking the current role

* Designed and implemented solutions to enterprise-level data unification problems at scale using in-house ML algorithms and third-party tools. * Consulted on best practices of materializing and/or categorizing entities (contacts, customers, suppliers, parts, registrations, etc.) to global and domestic enterprise customers. * Productionized a real-time notification module to monitor Spark jobs using RESTful APIs and Slack Web Client. * Delivered time-constrained POV projects leading to customer success. * Prototyped add-on applications to software that expands analytical functionalities.

* Built a data product called APRODUCE, a web application that uses machine learning to help online Peer-To-Peer (P2P) lenders to predict and track the revenue on their lending investment. * Ingested 2.5 million historical loan data with 140 features from lendingclub.com loan data repository. * Used logistic regression (scikit-learn) to classify loans that will default or not, resulting in a good balance with precision and recall metrics. * Used XGBOOST to obtain accurate predictions (1.6% MAE) on annualized percentage returns of any recommended (healthy) loans. * Deployed a web-based dashboard using Dash by Plotly and AWS Elasticbeanstalks. Provided easy-to-use interfaces and batch recommendations for both prospective and current loan investors