Ryze Wu ยท AI Agent Engineering
I build agents and the systems behind them.
I work across domain agents, tool use, workflow design, runtimes, routing, orchestration, retrieval, evaluation, tracing, and delivery. I care about clear boundaries, observable execution, and systems that remain useful outside a demo.
I designed and developed an enterprise Agent SDK and Runtime from 0 to 1.
I built a multi-Agent planning and orchestration control plane for complex goal-oriented tasks.
I developed configurable template agents for enterprise employee services and general customer support.
I standardized production engineering for Agent systems through evaluation, observability, build pipelines, and delivery.
Selected projects
View all projects- PleiadPublicI am building an open-source agent runtime for capability discovery, task graphs, and reliable execution.
- selfsigilPublicI built a local-first tool that turns Codex session evidence into capability profiles, skill maps, and resume drafts.
- PassionPublicI wrote an open-source behavior Skill for making LLM Agents more reliable under pressure.
Notes
Read the notesI write about the boundaries that make agent systems easier to build and operate: capabilities, context, runtime execution, orchestration, observability, evaluation, and delivery.