AI Application Engineer · Agent Systems · Verifiable Software Delivery

I build AI-native software where orchestration, memory, retrieval, tools, and interface quality work as one system. My focus is not code generation alone, but dependable behavior: explicit constraints, observable workflows, reproducible validation, and maintainable delivery.

Direction before decoration.

An Agent Skill orchestrator for premium brand websites, product launches, and cinematic Three.js experiences. It combines art direction, production frontend design, anti-template taste review, and selective WebGL specialists into one evidence-driven workflow.

View the workflow · Install the skill · Download v1.0.0

• ai-generate-code — Evidence-driven AI software synthesis and verifiable artifact delivery.
• memory-expand-windows — Experiments in durable memory and context expansion.

• Agentic systems: tool orchestration, layered memory, hybrid retrieval, and evaluation.
• Java engineering: concurrency, caching, streaming, state consistency, LangChain4j, and LangGraph4j.
• Interface systems: frontend direction, generative UI, WebGL, and interaction quality.
• Delivery discipline: builds, browser evidence, reproducible experiments, and engineering records.

I use AI to accelerate implementation while retaining ownership of the constraints, verification, and final behavior.