•Designing AI platforms, include machine learning architecture and data engineering ecosystem from scratch. Collaborate with stakeholders to understand business needs and translate them into actionable data science projects. Build a user-friendly customer engagement platform and mobile app. Attracting 30+ universities to join this platform.•Developing and transforming data science prototypes from model data preparation to model training to inferencing for complex and large-scale applications, include schema design, data collection and ingestion, data storage, distributed systems, feature engineering, model training and validation, integrating the latest advancements about recommendation method, Generative AI, LLM, NLP and prediction algorithm into our ML ecosystem.•Building end-to-end data customized pipelines. I independently own full stack handle programming technology completed customer data ETL Pipeline, translate model requirements into efficient data pipelines, ensuring optimal data quality, data clearing and analysis, processing, and integration. Successfully supported data driven platforms running.•Deployment of the model to AI backend as REST API server at production environment, realized container-based deployments utilizing containerization (Docker) and orchestration (Kubernetes) for high availability and scalability at cloud infrastructure. Moved from an offline environment and integrated into an existing production environment.•Collaboratively throughout the product lifecycle with software engineers, project manager and DevOps team by CI/CD pipeline and MLflow, Github, and kubernetes, using MLOps workflow continuously update data, code, and models, accomplish implementation of best practices for model versioning, experiment tracking, and model evaluation to ensure reproducibility and transparency, for seamless model deployment, monitoring, and ongoing optimization.

•Design industry-leading algorithms and platform using database, machine learning, deep learning, LLM to significantly distilling customer requirements into problem definitions and identifying measures of success.•Build recommendation system that includes collaborative filtering, content and model based, reinforcement learning, ranking algorithm, improve customer’s personalized recommendations experiences. Users click-through rate increased 15% and conversion rate increased 8%.•Implemented more than 10 machine learning, CNN, RNN, NLP and LLM models, analyzed the trend of time-series data, regression and classification, improved performance 10% by business meaning from data based on cross-validation, A/B testing.•Carried out churn and risk prediction model in analyze business operation data, analysis customer behavior features by XGBoost and LIME model, achieved 10% user volume includes 80% of lost users, saving 40% of resource costs.•One of the three scientists selected lead comprehensive knowledge of the data science training program for 200 engineers increasing measured unit proficiency by 15%.

•Organized 10 national-level research projects as the project leader, completed 15 national industry standard specifications, won 5 national-level awards, received the title of National youth outstanding young scientist.•Integrated more than 20 data sources, formulated data architecture and algorisms, hands-on theoretically programming 50 machine learning, deep learning, NLP and LLM models. •In-depth participation in model development, built more than 100 models of mathematic components, applied for recognition, classification and prediction. Calculated nearly 30 data products for decision support.•One of Top 1% National training teachers training program for 2,000 engineers. Presented more than 200 data analysis reports and presentations at the National Conference.