ERP/OS/Business-Finance Integration/Inventory and Logistics Management SaaS

ERP/OS/Business-Finance Integration/Inventory and Logistics Management SaaSs

Guangdong XXXXX Logistics Co., Ltd. (Feb 2021 - Jul 2024)

Position: Full-stack Engineer - Engineering & Architecture Team
Spearheaded the frontend refactoring of an ERP system (infrastructure, redevelopment, and deployment). Addressed legacy architecture complexities, high concurrency, and cross-department integrations in a containerized, disaster-tolerant deployment environment.

want know more?

ERP/OS/Business-Finance Integration/Inventory and Logistics Management SaaS

Redux ant-design Apollo Cluster

React + Nest.js + Tailwind CSS / Emotion designed for a multi-device scenario.Created map-rendering, distance location, and truck navigation components to address storage state persistence, employs the React-DAD visual design, Service workers were employed for offline caching in low-network environments, syncing to IndexedDB and submitting when online.integrates routing cost controls, a four-level address library for real-time automated settlement, flexible pricing Optimized initial rendering speed from 17s to 3s.
Employed AOP techniques using middleware, interceptors, guards, and pipes, combined with RBAC from JWT tokens, to dynamically determine user permissions. This solution decoupled access control, logging, caching, and data validation for centralized management.
Designed a BFF layer to streamline complex data flows in legacy systems. Leveraged GraphQL with Apollo Client's to unify multiple microservices APIs, reducing redundant requests and improving data query efficiency. Addressed high-concurrency N+1 query issues with DataLoader for batch data requests and utilized Cluster Module with multiple child processes, Redis Stream, and Pub/Sub for a caching layer to avoid frequent polling-induced slowdowns.
Enabled real-time communication and visibility of vehicle logistics through WebSocket on GPS location, speed, with a heartbeat reconnection mechanism and optimized scheduling algorithms for dynamic route updates, ensuring optimal load and efficiency for multi-million-tonnage logistics.
Monitoring and recovery: Sentry.io for logging and monitoring, capturing exceptions and performance bottlenecks. Kubernetes disaster recovery solution ensures quick restoration during unexpected failures, providing containerization for high availability. CI/CD combined with Jest and e2e testing.

AI Prompt-based Business Data Visualization & Analysis

Next.js GSAP prompt Echarts

Managed the project from inception to deployment, covering both frontend and backend development of a chart platform. This system utilized AI Prompts to generate strategic analysis charts based on input data sets and business analysis requirements, significantly improving the efficiency of data visualization.
Applied SSR in Next.js to enhance initial page load speeds, utilizing React.lazy and Suspense for code splitting and optimization. Memoized functions with UseMemo and improved the user experience through preloading techniques and GSAP-powered animations.resolved hydration errors during rendering.
Implemented virtual scrolling for loading large datasets and encapsulated custom hooks for chart rendering and pop-up detail windows. Integrated RAG to enhance prompt responses when passing data to LLMs by detecting question similarity using vector databases.