|
|
|
🧠 Software Engineering undergraduate focused on AI Application Development and AI Agents, actively learning AI Agent development and exploring AI applications in healthcare, education, and technology.
我是软件工程专业在读本科生,主攻AI 应用开发与AI 智能体方向,目前坚持学习 AI Agent 相关开发, 日常将实战代码片段整理至个人 Gist,持续探索 AI 在民生、教育、科技等场景的落地路径。
- 🔭 Focused on LLM App Development & Multimodal AI Systems · 目前专注于 AI 大模型应用开发 与 多模态智能系统 方向
- 🌱 Learning Python Backend & Multi-Agent Systems · 正在深入学习 Python 后端框架 与 AI Agent 多智能体开发
- 🔮 Exploring Embedded Systems & Intelligent Robotics · 未来计划拓展嵌入式开发与智能数字机器人的落地研发方向
- 🏆 Top 1% in Major, Multiple Scholarships · 多次获得专业奖学金,连续学年 专业排名前 1%
- 🥉 Regional 3rd Prize in Service Outsourcing Contest · Challenge Cup 3rd Prize · 服务外包创新创业大赛 东部赛区三等奖 · 挑战杯 三等奖
- 💡 Proficient in AI Dev Tools & LLMs, with full-cycle project delivery experience · 熟练使用多款主流 AI 编程工具与大语言模型,具备从需求到交付的全流程实践经验
- 📌 Goal: Become a Technical Architect who defines products from 0 to 1 · 目标:成为从 0 到 1 定义产品的技术架构师
日常使用的 AI 编程工具与个人智能体。
用于应用开发、方案探索与日常创作的大模型。
2026 年 AI 青年国际黑客松 · 北京线下 3 天开发 · 文化赛道
Interactive cultural travel platform built with React + JavaScript + Vite, featuring NPC-guided exploration, branching dialogues, story archives and journey collections. Set in Beijing’s Summer Palace, it offers illustrated storytelling and real-world exploration experiences, with continued iteration after the three-day on-site hackathon.
基于 React + JavaScript + Vite 构建文化旅行互动探索平台,以北京颐和园为首发场景,结合 NPC 同行、分支对话、故事档案与旅程图鉴,提供插画叙事与实景探索两套体验,实现观察、解谜与记录相结合的响应式 Web 交互,并在北京线下三天黑客松开发后持续迭代。查看源码 ↗ · 在线体验 ↗
校企合作项目 · 核心开发成员
SaaS multi-tenant beauty salon platform built with Ruoyi + FastAPI + Vue3 + Redis, featuring async task scheduling via Redis Stream, multi-level caching architecture, and AI-driven development workflow with Harness CI/CD automation.
基于 Ruoyi + FastAPI + Vue3 + Redis 构建 SaaS 多租户美容门店管理平台,采用 Redis Stream 实现异步任务调度,多级缓存 + 防雪崩架构大幅降低 DB 查询压力,落地 SpecCrew AI 驱动开发流程与 Harness CI/CD 工具,覆盖 5 类核心管理场景,预估预约处理效率提升 80%。
第十七届服务外包大赛 · 负责人 · 东部赛区三等奖
Multimodal emotional companion digital human powered by Qwen2.5-14B + FunASR + CosyVoice + MediaPipe + Live2D, with psychology RAG and real-time lip-sync, achieving 85%+ emotion recognition accuracy with <200ms latency.
基于 Qwen2.5-14B + FunASR + CosyVoice + MediaPipe + Live2D 构建“感知-认知-干预”主动闭环系统,采用多模态融合情感引擎与心理学知识库 RAG,实现语音识别→情感对话→数字人实时驱动的完整链路,情绪识别准确率超 85%,响应延迟低于 200ms,覆盖 8 类核心陪护场景。
2026年“挑战杯”参赛项目 · 核心成员 · 三等奖
Smart campus AI digital human platform built with LiveTalking + LangGraph ReAct Agent + FAISS RAG, supporting real-time WebRTC interaction, multi-model QA and domain knowledge retrieval across 3 scenarios.
基于 LiveTalking + LangGraph ReAct Agent + FAISS RAG 三大模块构建智慧校园 AI 数字人平台,支持 WebRTC 实时音视频交互、多模型智能问答与专有知识库检索,覆盖政务咨询、生活服务、学业指导三大场景。
XEngineer 新工科竞赛参赛项目 · 负责人
AI-powered spoken English training platform with real-time voice interaction, pronunciation scoring, grammar correction and quantified learning reports across 3 scenarios, using LangGraph dual-Agent workflow.
基于大模型 + 流式 ASR 构建 AI 英语口语陪练平台,支持面试/点餐/会议三类场景实时语音对话、发音评测、语法纠错与课后量化报告,采用 LangGraph 双 Agent 工作流。
浏览我的开源仓库 ↗ · 查看 Gist 开发笔记 ↗ · 访问个人网站 ↗
| 方向 | 技术与方法 | 实践场景 |
|---|---|---|
| AI 应用与智能体 | Prompt Engineering · LangChain · LangGraph · RAG · FAISS | 知识问答、Agent 工作流、多智能体协作 |
| 多模态交互 | ASR / TTS · FunASR · CosyVoice · Live2D · WebRTC | 数字人、情感陪护、语音陪练 |
| 后端与数据 | Python · FastAPI · Flask · Django · MySQL · Redis | API 开发、多租户系统、缓存与异步任务 |
| 前端与体验 | JavaScript · TypeScript · React · Vue · Next.js | 响应式网页、交互原型与产品界面 |
| 工程与交付 | Git / PR · Docker · CI/CD · 需求分析 · 技术文档 | 从需求梳理、开发协作到部署交付 |
| 持续学习 | C · Go(学习中) · 嵌入式与机器人(探索方向) | 拓展软件与智能硬件的连接 |
| 证书与语言 | CET-4 · CET-6 · 普通话二级乙等 · 全国计算机二级 | 英文资料阅读、沟通与表达 |





