Xiaomeng Li Email & Phone Number

California, United States

Redmond, Washington, Us

Work on Customer Experience Platform engineering team at Microsoft Azure- Designed and lead a C#, .NET, Azure SQL service for creating anomaly detection emails.- Designed and implemented a service for sending batch emails to subscribers weekly.- Spearheaded a cross-functional collaboration, partnering with managers and PMs.- Learnt and utilized Figma to design and finalize email UI templates in collaboration with PMs.- Built UI containers in React, TypeScript including shells, drop-downs, and pivots.- Created a complete UI table with different columns, groups supporting different table attributes.- Increased team micro services’ data reading geo-redundancy by 100% to Azure Data Explorer.- Created and published a NuGet package library for wrapping Azure Metrics Advisor SDK.- Set up Entity Framework for service in order to develop data models in Azure SQL.

Austin, Texas, Us

Worked on the Fusion Apps Control Plane team at Oracle Cloud Infrastructure- Refactored team’s Terraform codebase and built a new production pipeline.- Applied automated Terraform images, policies and resources.- Accelerated region rollout timelines from three months to two weeks, enhancing efficiency.- Region build as well as end to end tested all 26 commercial production regions for GA release.

Seattle, Wa, Us

Worked on the Business Web Services team of Amazon Business Org- Implemented new features for team’s Product Search Service in Java Spring.- Added integration tests and reached pipeline fully continuous delivery(CD).- Used AWS Key Management Service(KMS) SDK for encrypting different services’ logs.- Setup a complete AWS workflow configuring Kinesis for streaming logs and S3 for storing logs.- Configured AWS Redshift to subscribe to daily service logs and display business metrics.- The new workflow reduced PMs’ time to observe business metrics from weeks to same day.- Applied AWS DynamoDB to enable storing and querying user’s subscription settings.

Redwood City, Ca, Us

Worked on Natural Language Processing and Web Apps development- Developed a sentiment analysis pipeline to analyze customer and service agent conversations, leveraging Natural Language Processing (NLP) techniques to improve player experience within EA IT.- Implemented Google's Bidirectional Encoder Representations from Transformers (BERT) model using Python and TensorFlow on AWS EC2 instances with GPU acceleration, achieving 86% accuracy in sentiment classification on both customer conversation data (translated from audio using AWS Transcribe) and the IMDB movie review dataset.- Engineered a full-stack web application using Django, JavaScript, and Vue.js to provide customer support advisors with real-time access to sentiment analysis results stored in MongoDB. Integrated Google Charts for data visualization and implemented a news posting system for internal communication.- Deployed the application on AWS EC2 using Nginx as a reverse proxy, configuring a backend architecture that asynchronously applied the BERT model via Celery to process uploaded audio/text files and provide real-time sentiment predictions. Integrated AWS Transcribe directly into the web application to automate the audio-to-text conversion process.

Los Alamos, Nm, Us

Worked on time series forecasting with Deep Learning- Developed machine learning solutions to detect hidden seismicity on Mars as part of the Applied Machine Learning (AML) Summer Research Fellowship.- Preprocessed Marsquake waveform data using a Fingerprinting technique, transforming the results into both sequence and image representations to enable deep learning analysis from multiple perspectives.- Designed and implemented a Convolutional Recurrent Neural Network (CRNN) using Keras and TensorFlow, combining Convolutional Neural Networks (CNNs) for spatial feature extraction and Long Short-Term Memory (LSTM) networks for temporal sequence modeling. Optimized the CRNN architecture and hyperparameters to improve the detection of Marsquake events within time-windowed waveform segments used for training data. Evaluated model performance using 10-fold cross-validation and visualized results with confusion matrices. - Classified both waveform and Fingerprinting data using the developed CRNN and a Random Forest model, achieving a stable 80% accuracy on both training and testing datasets, demonstrating effective generalization and minimal overfitting.- Presented research findings at the AGU Fall Meeting: Larmat, C. S., & Li, X. (2018, December). Detecting the hidden seismicity of Mars. In AGU Fall Meeting Abstracts (Vol. 2018, pp. IN21D-0728).

Albuquerque, Nm, Us

Worked as a Ph.D. student- Published a journal paper.- TA’ed Engineering Statics (CE 202) class - TA’ed Structural Dynamics and Earthquake Engineering (CE 521) class.Courses Taken:CS 561 Algorithms & Data StructureCS 512 Introduction to Computer GraphicsCS 529 Introduction Machine LearningCS 544 Intro to CybersecurityCS 564 Intro Database ManagementCS 558 Software FoundationsCS 585 Computer NetworkCS 527 Principles of Artificially Intelligent MachinesECE 437 Computer Operating SystemsCS 251 Intermediate Programming With JavaCS 471 Intro Scientific ComputingCS 591 ST: Advanced Neural NetworksCS 650 Reading and Research

Kanagawa, Jp

Asia-Pacific-Euro Summer School on Smart Structures Technology (APESS 2017)

Baltimore, Md, Us

Worked at Thin-Walled Structures Lab- Built-up studs lab experiment work.Worked on uncertainty quantification and probabilistic computational mechanics- Python probabilistic simulationCourses Taken:560.604 Solid Mechanics for Structures560.618 Probabilistic Methods560.619 Advanced Structural Analysis560.730 Finite Element Methods560.601 Applied Math for Engineers560.348 Probability & Stats for Engineers550.420 Introduction to Probability560.630 Structural Dynamics

Master Of Science - Ms, Structural Engineering

Master Of Science - Ms, Computer Science

Doctor Of Philosophy - Phd, Civil Engineering (Dropped)

Bachelor Of Engineering - Be, Water Supply And Sewerage Engineering