Jie Liu

I am currently a Tech Lead Manager at SHEIN Technology LLC, the R&D-focused technology subsidiary of SHEIN Group, one of the world’s largest fast-fashion companies. I lead the Augmented Intelligence team within the US R&D organization, covering AI and Commerce Engineering initiatives, and report directly to the US CTO. I also lead AI agent and skill development initiatives for the Big Data Engineering team. My organization focuses on applying AI, computer vision, multimodal systems, and LLM technologies to modernize engineering productivity, enterprise operations, commerce platforms, and intelligent automation across SHEIN’s global ecosystem. With the rapid adoption of generative AI technologies, I led the standardization of AI coding assistants such as Claude Code and Codex across R&D workflows for both engineering and operational use cases. Our team has developed hundreds of internal AI skills and automation solutions to improve developer productivity, operational efficiency, and AI adoption at scale. The initiatives led by my team include BDE AI automation, AI-driven marketplace systems for GMV growth, PDA management platforms for global warehouse operations, the SHEIN Open/Developer Platform, enterprise emergency notification systems, and AI-powered compliance solutions supporting HR and legal organizations. In parallel, I lead cross-functional AI initiatives involving engineering, infrastructure, compliance, legal, product, and business teams to drive enterprise-scale deployment and governance of AI systems. I am responsible for technical strategy, organizational alignment, architecture decisions, and execution planning for large-scale production systems under aggressive timelines. I also play an active role in organizational growth and talent development across the R&D organization, including technical hiring, AI-focused interviewing, engineering mentorship, and leadership development initiatives.

Led the development and production deployment of enterprise AI systems for SHEIN, focusing on content governance, legal compliance, fraud detection, and multimodal intelligence platforms. In 2024, our team successfully launched large-scale product review and legal protection systems designed to support internal reviewers, suppliers, and legal teams in identifying potential infringement, prohibited content, copyright risks, and intellectual property conflicts across SHEIN’s ecosystem. The platforms integrated multimodal AI technologies, including OCR, image retrieval, LLM-based reasoning, and computer vision models, to improve operational efficiency, compliance automation, and risk management workflows. In parallel, I led fraud detection initiatives in collaboration with data compliance and risk-control teams to strengthen payment security and transaction risk analysis across SHEIN platforms. The machine learning models were deployed into production environments and evaluated through large-scale online experiments and A/B testing. I also collaborated closely with cross-functional teams across AI engineering, compliance, legal, infrastructure, and business operations to ensure scalable deployment, governance alignment, and enterprise adoption of AI-driven systems.

Selected as a member of Google Cloud’s Compute and AI Infrastructure Customer Council, collaborating with Google leadership and industry peers on the future of AI infrastructure, cloud computing, large-scale machine learning systems, and enterprise AI adoption. Contribute strategic feedback and technical perspectives on AI platforms, generative AI infrastructure, developer productivity, scalable compute systems, and next-generation cloud technologies influencing the evolution of Google Cloud’s AI ecosystem.

Awards: • Technology Breakthrough Award, OPPO Research Institute (2022) • Outstanding Team Award, OPPO Research Institute (2022 H2) • Outstanding Team Award, OPPO Research Institute (2022 H1) • Outstanding Team Award, OPPO Research Institute (2021 H2) At OPPO U.S. Research Center (InnoPeak Technology, Inc.), I contributed to the development of core tracking and perception algorithms for XR and mixed reality systems. Our team developed advanced hand-tracking and controller-tracking solutions for AR/MR headsets by integrating multi-sensor fusion techniques across image sensors and IMU data streams to achieve accurate, low-latency, and real-time spatial tracking. I worked closely with cross-functional teams across hardware, industrial design, SDK, applications, and engineering validation to support the end-to-end development and productization of OPPO’s next-generation XR devices. In 2023, we successfully delivered and demonstrated OPPO’s mixed reality headset platform at AWE USA 2023, powered by Qualcomm XR2 technologies and proprietary tracking algorithms developed internally. In addition, I contributed to next-generation XR interaction systems, real-time perception pipelines, and foundational spatial computing technologies for immersive human-computer interaction experiences.

Awards: • Outstanding Team Award, OPPO Research Institute (2021 H1) • Outstanding Individual Award, OPPO Research Institute (2020 H2) • Breakthrough Innovation Award, OPPO Research Institute (2020 H1) • Outstanding Team Award, OPPO Research Institute (2019 H2) At OPPO U.S. Research Center (InnoPeak Technology, Inc.), I worked on large-scale computer vision and deep learning systems for face recognition, including face detection, alignment, and verification using modern neural network architectures. Later, I became one of the founding members of the hand tracking and recognition initiative for XR and spatial computing applications. Our team built the hand tracking group from scratch in 2019 and successfully delivered OPPO’s first internally developed hand tracking system for AR Glass products in 2020. I led and contributed to deep learning and optimization-based computer vision solutions for real-time hand tracking, gesture recognition, and human-computer interaction systems. The technologies were further extended to XR and IoT devices, including smart TVs and wearable platforms. In addition, I contributed to foundational AI algorithm development for next-generation interactive systems and participated in the submission of more than 20 related patent applications.

• Developed predictive models for liquidity analysis and financial market event forecasting using large-scale market datasets. • Researched and evaluated recurrent neural network architectures, including GRU and LSTM models, for financial time-series prediction tasks. • Designed and implemented convolutional neural network (CNN)–based forecasting approaches for sequential market data analysis. • Contributed to the adoption of deep learning methodologies within the team by supporting technical discussions, model implementation, and knowledge sharing initiatives.

• Conducted research on scalable second-order optimization algorithms for large-scale machine learning systems at Tencent AI Lab. • Developed accelerated L-BFGS optimization approaches by integrating mini-batch second-order approximations for efficient large-scale model training. • Performed theoretical convergence analysis and optimization validation for stochastic and distributed learning settings. • Proposed computationally efficient implementations of second-order optimization frameworks for practical deep learning applications. • Contributed to research published in a NeurIPS workshop.

• Conducted research on anomaly detection and predictive analytics for large-scale manufacturing systems at Mitsubishi Electric Research Laboratories (MERL). • Developed deep learning–based anomaly detection frameworks using Autoencoder architectures for industrial sensor reconstruction and fault detection. • Designed structured neural network models tailored to real-world manufacturing production lines using sequential sensor data and domain-specific feature engineering. • Evaluated and benchmarked anomaly detection approaches including Support Vector Machines (SVM), Isolation Forest, Local Outlier Factor (LOF), and neural-network-based forecasting models. • Contributed to research publications presented at the IEEE World Congress on Intelligent Control and Automation (WCICA) and related USPTO patent work.

• Conducted research on large-scale optimization problems in smart grid and power systems under the EU Horizon 2020 initiative at IBM Research Ireland. • Developed scalable optimization approaches for Alternating-Current Optimal Power Flow (AC-OPF) problems by combining first-order randomized optimization techniques with second-order Newton-based methods for improved convergence efficiency. • Applied advanced mathematical optimization and semidefinite programming (SDP) relaxation techniques to large-scale energy system modeling and real-time power flow optimization. • Extended optimization frameworks to renewable energy systems with time-varying and real-time operational constraints. • Contributed to research publications in IEEE Transactions on Smart Grid and the IEEE Power Systems Computation Conference (PSCC).

• Conducted research on Gaussian Process modeling for energy-related time-series forecasting and probabilistic predictive analytics. • Developed predictive maintenance models for turbine event forecasting and operational analysis in wind farm systems. • Designed and implemented interactive forecasting and visualization tools for customer-facing energy analytics applications using R Shiny. • Applied statistical learning and time-series modeling techniques to support industrial energy optimization and maintenance decision-making.

• Conducted research on time-series analysis and predictive modeling of power plant energy data within the Machine Learning Group at Siemens Corporate Research. • Developed statistical forecasting models based on ARIMA and wavelet transform methodologies for large-scale energy consumption prediction. • Designed and evaluated deep learning approaches for time-series forecasting, including Feedforward Neural Networks (FNNs) and Recurrent Neural Networks (RNNs) such as Elman and Jordan architectures. • Implemented and validated statistical and neural forecasting models in R, with a focus on predictive accuracy, temporal pattern modeling, and industrial energy analytics.

• Conducted research on regression modeling and energy-efficiency optimization within the Mathematics and Computer Science (MCS) Division at Argonne National Laboratory. • Developed advanced sampling methodologies based on k-means clustering and Latin hypercube sampling for high-dimensional statistical modeling. • Designed and implemented Bayesian Gaussian Process regression frameworks with multiple kernel functions for nonlinear predictive modeling and uncertainty estimation. • Developed research software and experimental prototypes in R and MATLAB to evaluate statistical performance and regression accuracy across different modeling approaches.