| Pengbin Li | AI Agent Engineer |
AI Agent Engineer · Applied AI Researcher
Master's student at Iwate University working at the intersection of AI agents, retrieval-augmented generation, cybersecurity, and deep learning. I move from problem framing and experiments to backend implementation, evaluation, observability, and production-oriented delivery.
View selected projects Read full CV Contact me
Selected impact
- 116 · real agent evaluation cases
126 turns; 100% pass across the latest full evaluation - 93.3% · RAG Top-1 accuracy
up from 53.3%; Top-K reached 100% - 98.7% · first-feedback latency reduction
0.4145s → 0.0052s in a slow-search simulation - 96★ · open-source validation
Paper Humanizer Skill on GitHub at CV publication time
What I build
01 · Agent systems
Design stateful single- and multi-agent workflows with planning, memory, tool routing, reflection, human handoff, and long-running task control.
Keywords： LangGraph · LangChain · Deep Agents · Harness · Multi-Agent
02 · RAG quality & evaluation
Build retrieval pipelines with query rewriting, filtering, reranking, evidence sufficiency checks, citations, offline metrics, trace-based regression, and graceful fallback paths.
Keywords： ChromaDB · pgvector · Elasticsearch · Neo4j · Golden Sets
03 · AI × security
Apply LLMs and generative models to authorized security assessment, intrusion detection, vulnerability research, guardrails, sandboxing, and auditable execution.
Keywords： Security Agents · IDS · RAG · Docker · Observability
View the full capability matrix
Selected projects
Six projects that show how I approach agent orchestration, RAG quality, safety, evaluation, and applied deep learning.
Trip Planner · AI Travel Planning System
Jun 2026
A Chinese-language AI travel planning product that connects fast itinerary generation, deep destination research, conversational editing, maps and weather enrichment, history management, and Markdown/PDF export into one continuous workflow.
Stack： Vue 3 · TypeScript · FastAPI · LangGraph · ChromaDB · Redis · SQLite · RAG
Enterprise Java AI Agent Platform Enhancement
Jun – Jul 2026
Extended java-up-up/super-agent with production-oriented capabilities for tool extensibility, RAG evaluation, retrieval closure, security, observability, and citation traceability.
Stack： Java · Spring AI Alibaba · RAG · Neo4j · Redis · Kafka · pgvector · Elasticsearch
MyAgent · Defensive Security Assessment Agent
May – Jun 2026
A PI Agent-based CLI/TUI system enhanced with authorized defensive-security subagents, scoped execution, security memory, multi-stage planning, and report generation.
Stack： TypeScript · Node.js · PI Agent · CLI/TUI · Multi-Agent · Security Guardrails · Vitest
Paper Humanizer Skill
Jan 2026
An open-source bilingual academic-writing tool that reduces formulaic AI writing patterns while preserving facts, numbers, citations, conclusions, and domain terminology.
Stack： Claude Code Skill · Python CLI · Prompt Engineering
Explore all project case studies
Experience
Java Development Engineer · Hunan Kechuang Information Technology Co., Ltd.
Jul 2023 – Nov 2023
Contributed to Changsha's municipal procurement system with Java, Spring Boot, and MySQL. Implemented backend APIs and data-processing logic, including multi-format tender-document parsing, key-information extraction, and search; also participated in requirements analysis, testing, and defect resolution.
Education
M.S. candidate, Design and Media Engineering · Iwate University · Japan
Oct 2024 – Present
- University scholarship · GPA 3.2/4.0
- Coursework: computer vision, integrated design, computer network systems, computer animation, design representation, image synthesis
- Research: GAN-based intrusion detection using synthetic and real traffic data to improve generalization and adversarial robustness
B.Sc., Information and Computing Science · Central South University of Forestry and Technology · China
Sep 2019 – Jun 2023
- GPA 3.0/5.0
- Coursework: Java/C/C++/Python, MySQL, computer organization, operating systems, algorithms, data structures, information security, data visualization, software engineering
How I work
I am problem-driven: I decompose ambiguous goals, define measurable checkpoints, and iterate from a working baseline toward a reliable system.
I enjoy using new development paradigms and AI coding tools to validate ideas quickly, but I care equally about tests, evidence, safety boundaries, and maintainable engineering.
Basketball has reinforced how I think about collaboration: clear roles, fast feedback, shared context, and sustained execution matter as much as individual skill.
Contact
- Email： cralpbin@gmail.com
- GitHub： github.com/crabin
- Blog： cnblogs.com/crabin