陈韬 Chen

In Work‑Space, I explored text processing, content generation, and question-answering scenarios, designing structured prompts and interaction flows. I tested model limitations and documented validation rules and fallback approaches.

For market‑workbench, I researched business users’ needs and planned a lightweight AI workspace for information analysis, summaries, and content generation. I mapped user workflows and considered implementation costs and risks.

On yiji, I plan Agent workflows around task planning, tool use, memory, and reflection. I define interaction patterns, failure fallbacks, and task-ending conditions, then update the workflow based on testing feedback.

In my AI video generation product practice, I explored text-to-video and image-to-video workflows. I planned controls for camera shots, style, motion, and duration, and defined acceptance criteria for issues such as flicker, distortion, and camera shake.