Wayturn Hung （洪偉騰）

I’m Wayturn Hung （洪偉騰）, a backend engineer in Taiwan focused on applied AI and enterprise modernization. Backend engineering is the foundation of my work: clear contracts, observable behavior, and careful changes to systems people rely on.

My current work connects three areas: planning behavior-preserving changes to existing systems, building AI-assisted engineering workflows, and evaluating what those workflows keep or change.

My backend foundation is primarily C#/.NET, ASP.NET Core, SQL, and Docker. My public Applied AI projects use Python to explore LLM integration, developer tooling, evaluation, and inspectable AI-assisted workflows. They extend the same backend practice of making system behavior understandable and verifiable.

Enterprise modernization

I am responsible for technical discovery and proof-of-concept planning for an enterprise modernization effort. The public case shares its technology context and general method: understand existing behavior, choose a bounded example, and plan how to compare results before considering a broader migration. The work is in progress; the migration is not complete.

Applied AI and developer tools

DevAgent CLI makes a small AI-assisted development workflow traceable. Jev Context Prunertests model-assisted conversation pruning with fixed safeguards, full-input fallback, traces, and evidence-based evaluation. Both remain bounded projects, with their methods and limits described alongside the results.

What I am exploring next

I want to keep applying backend engineering to AI systems and developer platforms, with particular interest in work that brings models into real applications while keeping inputs, decisions, failures, and verification understandable.

This site records the work and its evidence. The AI Visibility Lab is a separate research track about how the public record is understood; its historical baseline is not a measure of this positioning update.

B.S. in Earth Sciences, National Taiwan Normal University.