Daniel Carter
AI Application Engineer · 5 years of agent engineering
+1-555-0142 · daniel.carter@example.com · Austin, TX
GitHub github.com/example · Blog blog.example.com
Summary
5 years of full-stack web and AI application experience. Led 3 agent product lines from zero to first release, serving over 200,000 users. Turns model capabilities into testable engineering modules, with solid practice in tool-call reliability, context cost control, and eval loops.
Work Experience
- Owned retrieval and tool-call pipelines for an enterprise knowledge-base agent, lifting first-answer accuracy from 61% to 84%.
- Built a 420-case prompt regression suite with automated scoring, cutting post-upgrade regressions from 12 to under 2 per week.
- Reworked multi-turn context compression, reducing token cost by 43% while holding task completion steady.
- Led the analytics dashboard migration from Vue 2 to React 18 + TypeScript, cutting build time by 55%.
- Built a component library reused by 4 product lines, shrinking average new-page effort from 3 days to 1.
Projects
github.com/example/agent-runtime
- TypeScript multi-channel agent runtime covering session isolation, skill loading, and long-term memory.
- Markdown-defined skills loaded on demand, cutting per-turn context cost by about 60%.
Education
- Data structures, operating systems, databases, intro to machine learning. GPA 3.6/4.0.