Curriculum Vitae


Education

Ph.D. in Software Engineering, Minor in ECE
University of Arizona
Advisors: Prof. Bo Liu (RL Turing Award Lab) and Prof. Janet Roveda (BME).
Research focus: AI Safety, Alignment, Reinforcement Learning, Agentic AI, AI for Science & Healthcare.
M.S. in Economics and Computer Science; Master of Management Studies
Duke University
GPA: 3.8/4.0 (Top 10% Fuqua Scholar)
Bachelor of Science (Hons)
Oxford Brookes University

Work Experience

AI Engineer III, OPPO US Research Center
  • Led AI evaluation research for large language model and multimodal agents, designing benchmarking frameworks and end-to-end evaluation pipelines across text, image, video, and audio generation tasks, while researching long-horizon memory, context engineering, and agent workflow optimization.
  • Optimized LLM inference infrastructure for Qwen models with vLLM and Kubernetes, applying quantization, tensor and pipeline parallelism, and advanced serving techniques to improve GPU utilization, throughput, and end-to-end latency at production scale.
  • Researched AI agent safety and reliability through systematic red teaming, adversarial testing, and edge-case evaluation, and developed guardrails, runtime monitoring, and risk assessment frameworks that turn findings into trustworthy production safeguards.
Head of Research Engineer, Anyidea Inc
  • Designed and shipped retrieval-augmented (RAG) AI agent workflows in Python with Flask/FastAPI and LLM APIs, integrating retrieval pipelines, tool use, and backend services to automate multi-step SaaS business processes for enterprise applications.
  • Built AI system monitoring and analytics infrastructure with dashboards, structured logging, and alerting pipelines to track agent performance, workflow execution, latency, and reliability metrics across services.
Strategist, Perplexity
  • Promoted Perplexity's AI agent products through technical demos and community outreach, showcasing Deep Research, multi-model capabilities, Spaces, and the Comet AI browser for AI-powered research and productivity workflows.
Research Intern, Frost & Sullivan – LeadLeo Research Institute
  • Researched LLM evaluation methodologies and built automated benchmarking pipelines to measure model capability, robustness, and reliability across finance and healthcare domains, surfacing systematic failure modes.
  • Developed RLHF-based reward modeling and alignment frameworks to optimize LLM behavior toward safe, reliable, and domain-appropriate responses for real-world deployment.
Lead Research Engineer, Srlon
  • Built large-scale data infrastructure with Kafka-based telemetry pipelines for AI-powered e-commerce applications, enabling reliable user data ingestion, real-time analytics, and downstream machine learning workflows.
  • Developed cloud data platforms using Datastream (CDC) and BigQuery to analyze customer monetization data and power KPI dashboards that guided product and business decisions.

Selected Publications & Projects

Publications

Loading publications…

Projects

Improved Regularization for Convolutional Neural Networks
Xinyi Xie, Weisheng Jin, Chen Dong
Symbolic Knowledge Distillation
Yuchen Cao, Xinyi Xie, Mason Ma, Nathan Nguyen, Dennis Tang

Technical Skills

  • AI Technical Skills: Post-training, alignment, reinforcement learning, supervised fine-tuning, reasoning, multimodal, evaluation
  • Programming & Development: Expert in Python, Java, JavaScript/TypeScript, HTML/CSS, SQL; Proficient in Flask, FastAPI, Node.js, React.js, RESTful, Swagger
  • Database & Cloud Platforms: Skilled in PostgreSQL, MySQL, MongoDB, Hadoop/Hive, DynamoDB, Redis, VectorDB; Experienced with AWS (S3, EC2, EMR, Lambda, EKS, SQS, SNS), GCP (Compute Engine, App Engine), Azure
  • Big Data & DevOps Tools: PySpark, Kafka, PyTorch; Containerization with Docker/Kubernetes; CI/CD with Jenkins/Git; Infrastructure as Code with Terraform; Test & Monitor with JUnit/Postman/Grafana/Prometheus

Miscellaneous

Hobbies: Golf, Yoga, Pilates, Nutrition, and Trying all kinds of food

Teaching Experience

Main Teaching Assistant, Duke University — Quantitative Business Analysis
  • Led weekend review sessions and guided students on probability, statistics, and AI/Deep Learning applications; designed and graded exams and quizzes, and provided support for TA sessions.
Teaching Assistant, Duke University — Applied Probability and Statistics