AI Engineer

Building tools that make small teams unfairly fast.

I'm Daniel Ko, an engineer working on applied LLM systems, evaluation workflows, and product tooling that help teams ship faster without losing trust.

Selected Papers & Talks

Practical Retrieval UX for Team Knowledge Systems, 2025
Making Guardrails Visible to Product Teams, 2024
Evaluation Loops for Internal AI Assistants, Applied AI Forum

What I Work On

Prompt systems, tool orchestration, lightweight evals, internal copilots, and the product surfaces that make AI output usable in day-to-day work.

GitHub Repositories

agent-evals-kit

Reusable scoring flows for prompt and toolchain regression checks.

support-summarizer

A triage assistant for support threads with escalation signals and memory prompts.

r2-static-pipeline

Deployment helpers for static AI-generated sites with verification and rollback notes.