• Building AI copilot for traditional audits - streamlining cash tie-out and Monetary Unit Sampling • Shifting audit workflows from experience-based to AI-driven accuracy and speed, reducing manual audit work from hours to minutes through intelligent automation • Digging deep into Ethereum, Bitcoin, and stablecoin ecosystems, offering transaction tracing, wallet verification, and smart contract analysis — pushing on-chain financial integrity forward.

• Actively participated in an intensive 8-week pre-acceleration program in Palo Alto, California. • Engaging with Silicon Valley's top minds, venture capitalists, and innovative founders, enhancing my skills in pitching, business development, and networking to prepare for the startup ecosystem.

• Building a data pipeline for internal use, processing 1TB+/day with ~50ms latency. • Cut analysis time by 60% and supported 10K+ concurrent users with Redis caching and ECharts dashboards.

• Built a distributed prompt-management & real-time sentiment analytics platform (Spark + Kafka + Redis + Elasticsearch), processing 1TB+/day with ~50ms latency. • Cut analysis time by 60% and supported 10K+ concurrent users with Redis caching and ECharts dashboards.

• Designed a Spring Cloud microservices learning platform (Eureka, Gateway) with WebSocket clusters + Redis for 100+ concurrent users. • Built a Docker/K8s code-execution engine with RabbitMQ queues, handling ≈500 concurrent submissions at ~50ms latency.

• Built a high-performance data pipeline (Python, Dask, Arrow, Hadoop) with lazy loading + bucket partitioning, doubling processing speed. • Enhanced model accuracy 40% and reduced false negatives 30% with GPU acceleration (RAPIDS/CUDA, Numba).