# Luke Cheng [LinkedIn](https://www.linkedin.com/in/luke-cheng/) | [GitHub](https://github.com/Luke-Cheng) ## Experience ### Software Engineer | JPMorgan Chase, New York, NY > Dec 2025 – Present - Contributed to the query side of a CQRS-style centralized customer data platform, building the Customer Summary Utility to consolidate customer information from multiple internal systems and serve 10M+ daily requests (~10K TPS) with p99 latency below 200 ms using Spring Boot, Kafka, Cassandra, and CockroachDB. - Implemented a Change Data Capture design pattern to propagate database changes through lightweight Kafka events rather than transmitting full customer payloads, reducing messaging overhead and enabling lower-latency updates to the customer-read model. - Implemented reactive Kafka consumers with batch processing to asynchronously process customer-data changes at high throughput, keeping the query-side data model synchronized while reducing coupling between upstream systems and the customer-facing API. - Supported the platform's transition to Kubernetes-based infrastructure as traffic increased, improving horizontal scalability, deployment consistency, and operational reliability for high-throughput services. - Expanded automated testing across diverse customer datasets—including addresses, phone numbers, account-to-customer relationships, account identifiers, and card data—using JUnit and Cucumber, providing regression coverage for data-processing paths that could not be reliably exercised through UAT alone. - Extended the API and underlying data model to support multiple physical and virtual card numbers associated with a single account, enabling customer-data services to accommodate mobile-wallet and virtual-card use cases without relying on one-to-one account-to-card mappings. - Standardized deployment configuration across environments to improve consistency between test and production, reducing environment-specific configuration issues and increasing confidence in production releases. - Implemented Dynatrace SaaS observability to provide application and service-level visibility across the distributed platform, enabling developers to identify performance and health issues before diving into detailed Splunk logs. - Improved release safety through Helm-based Canary deployment and rollback strategies within Jenkins CI/CD pipelines, enabling controlled production releases and faster recovery from deployment issues. - Helped maintain operational continuity during a team transition by troubleshooting Kubernetes deployments, CI/CD workflows, configuration issues, and legacy service behavior while documenting system knowledge for continued development and support. ### Co-Founder | MOYU LLC, Pittsburgh, PA > Feb 2023–Jan 2024 - Founded an early-stage startup exploring semantic version control for CAD/BIM assets, addressing workflow limitations for complex engineering design files. - Developed cross-platform visualization and revision-diff tooling using Next.js, Electron, and Three.js for large-scale 2D and 3D architectural models. - Researched multi-module computer vision and geometric reasoning techniques for engineering design understanding to inform product strategy. ### Graduate Software Developer | University of Pittsburgh Medical Center, Pittsburgh, PA > Feb 2023–Jan 2024 - Built a HIPAA-compliant clinical analytics platform by translating requirements from 50+ clinicians into scalable healthcare software using Java, Spring Boot, MySQL, FHIR APIs, and AWS. - Developed Python NLP pipelines using spaCy, NLTK, and custom information extraction techniques to convert unstructured clinical notes into structured medical knowledge. - Designed semantic indexing and intelligent clinical search workflows to improve retrieval of medical information. - Deployed secure cloud infrastructure supporting high availability, regulatory compliance, and scalable healthcare data processing. ### Research Scientist | ChemPacific Corp, Baltimore, MD > May 2021–May 2022 - Applied data science techniques to laboratory manufacturing datasets, building Python workflows that improved process visibility and operational decision-making. - Designed automated laboratory data pipelines integrating Agilent analytical instrumentation with enterprise systems, improving traceability and reducing manual processing. - Performed statistical analysis, data validation, and process optimization across large experimental datasets supporting manufacturing quality and scientific research. - Automated laboratory workflows using Python and PowerShell, eliminating repetitive manual tasks and saving 150+ hours annually while maintaining GMP and NMPA compliance. ## Education ### Master of Science in Computer & Information Science | University of Pittsburgh > Aug 2022–May 2024 ### Bachelor of Science in Chemistry and Philosophy | Virginia Tech > Aug 2016–Dec 2020 ## Personal Projects ### Small-Form-Factor PC Checker | Next.js, Vercel, Google Cloud - Developed a web app and Chrome extension to verify SFF PC part compatibility using AI-driven manual parsing and Google Custom Search. ### Android Morse Code Keyboard | Android, Java, Kotlin, Jetpack Compose - Designed a custom Android IME to support Morse code learning with haptic feedback and accessible UI. ### Cryptography & Twofish Encryption | Java - Implemented cryptographic algorithms with an emphasis on security principles and clean object-oriented software design. ### Paint Sales Demand Forecasting | R, Python - Developed predictive machine learning models on PPG Industries datasets, applying feature engineering and sales forecasting to guide inventory planning. ### Screeps Doc/API Localization | Git, GitHub Actions - Led an open-source initiative to translate Screeps API documentation into Chinese (zh-CN), building automated GitHub Actions CI/CD workflows for validation. ## Leadership & Activities - **CyberForce Competition Team Lead**: Led University of Pittsburgh team, securing infrastructure and hardening defenses against simulated cyberattack scenarios. - **Virginia Tech Parkour Club Treasurer**: Managed club finances, secured 60% gym discounts for members, and navigated leadership transitions during COVID-19. - **Games4SocialImpact Winner**: Recognized with First Penguin Award for "Bon Voyage". - **Founder of Screeps China**: Spearheaded open-source API and documentation localization for the Chinese developer community. ## Interests & Hobbies - **Skiing**: Active meditation and outdoor focus. - **Philosophy**: Critical thinking, ethics, and epistemology. - **Gaming & Simulation**: - Automation & Systems: Screeps, Factorio (modular design, programmatic problem-solving). - Resource Management: Oxygen Not Included (prioritization and system balance under constraints). - Narrative & Ethics: This War of Mine, Papers, Please (high-stakes decision-making and ethical dilemmas). - Visual & Interactive Design: Return of the Obra Dinn, What Remains of Edith Finch (deduction, atmosphere, and visual storytelling).