At Aurelia Ng, I engineered a business intelligence search aggregator with Python, FastAPI, and Next.js, consolidating 10,000+ company records from more than 10 data sources. I designed probabilistic deduplication for noisy third-party data, reaching 80% precision while optimizing query latency to about 50ms.
At Halow Continental, I architected a webhook processing pipeline with Golang, Gin, and Redis queues. Async messaging, idempotency keys, and retry logic increased throughput 20x and cut failed deliveries by 90%.
Through Freelance & Enterprise Contracts, I architected a hotel booking SaaS deployed across 250+ properties and developed backend systems and smart contracts for a Polygon-based real estate tokenization platform. At Lead City University, I trained and benchmarked fake-news classification models and deployed an inference API with less than 50ms latency.

