一番下に日本語版もあります
AI Agent Researcher / Software Engineer — Bridging the gap between Art/Design (UI/UX) and Backend Logic.
As a university student, I specialize in building intuitive AI Agents that interact with the physical world and daily life. My background in Art & Design allows me to craft superior user experiences, while my technical expertise ensures robust system architecture.
Vision: I aim to implement AI agents in society to foster a world where human communication flourishes. By delegating daily tasks to AI, I want to create a society where people can truly focus on meaningful, face-to-face interactions.
- Focus: LLM Agents, Browser Agent, RAG, Edge AI, UI/UX Design.
- Location: Japan | Languages: Japanese (Native) / English
"Orchestrated a multi-agent system with an orchestrator and combined long-term/short-term memory, achieving a 1.7x improvement in task success rate."
Developed a Multi-Agent Orchestration System for my graduation research, integrating five specialized agents (RAG, Browser Use, IoT, Scheduler, and Orchestrator). This research was presented at NCSP'26 (Honolulu, Hawaii). The experience of integrating and evaluating a system that provides real-world feedback through hardware, rather than just remaining within the web, is a rare achievement in Japan.
- Resume: https://github.com/kota-kawa/resume
- Portfolio: https://project-kk.com/
- LinkedIn: https://www.linkedin.com/in/kota-kawa/
- Email: kota7kawagoe@gmail.com
Direct access to my currently active applications and platforms.
- [Yorozu-no-Madoguchi] (https://chat.project-kk.com/) — AI Concierge Window (Multi-Agent System).
- [ChatCore-AI] (https://chatcore-ai.com/) — Full-stack AI Chat Platform & Prompt Sharing.
- [FS-QR] (https://fs-qr.net/) — File Sharing & QR Code Generator.
-
Browser-Agent — Control the web with natural language.
- Overview: A browser automation agent with a FastAPI web interface. Executes tasks via natural language instructions with real-time monitoring via noVNC.
- Performance: WebArena Shopping task success rate: 32.6%.
- Technologies: Python, FastAPI, TypeScript, Docker.
-
ChatCore-AI (Live) — Production-ready AI chat platform.
- Overview: A full-stack chat system with email authentication, persistent conversations, and prompt sharing. Integrates Groq and Gemini.
- Technologies: Python (FastAPI), Next.js, PostgreSQL, Docker.
-
Symphony Agent Conductor — (Autonomous Orchestration Platform)
- Overview: An autonomous orchestration platform using LangGraph to coordinate specialized agents (Browser automation, IoT control, RAG) for seamless interaction between digital tasks and physical device control.
- Technologies: Python, Flask, LangGraph, browser-use, IoT (Jetson/Pi), OpenAI/Gemini.
- Keio University, Graduate School of Media and Governance
- Master's Program in Cyber Informatics
- Kanagawa, Japan | April 2026 – Present
- Kanagawa Institute of Technology (KAIT)
- Faculty of Information Technology
- Department of Information Network and Communication
- University of North Alabama — Exchange Program Completed
- Alabama, USA | May 2023 – May 2024
| Category | Technologies |
|---|---|
| Programming Languages | Python, TypeScript, SQL |
| Web / Application Frameworks | FastAPI, React, Next.js, Tailwind CSS, PostgreSQL, Redis |
| AI / LLM Engineering | RAG, Multi-Agent Systems, LangChain, LangGraph, browser-use |
| Infrastructure | Docker, AWS (EC2, VPC, Systems Manager), Linux, Nginx, Git, GitHub Actions, pytest |
| IoT / Hardware | NVIDIA Jetson Orin Nano, Raspberry Pi 4/Pico W |
- Building: A unified chat interface (inspired by OpenClaw & Hermes Agent) for browser-to-OS tasks on Ubuntu via natural language, with permission management and execution control for safe, practical agent deployment.
- Building: Improving the reliability and stability of AI agent execution in production environments.
Last updated: 2026-05-18
日本語版(Japanese Version)
AI Agent Researcher / Software Engineer — Art/Design (UI/UX) と バックエンドロジック。
大学でLLMを用いたAIエージェントの研究をしています。 Art/Design のバックグラウンドを持ち、優れた UI/UX と堅牢な バックエンドロジック の両方を兼ね備えたエージェント開発が得意です。
Vision: AIエージェントの社会実装を通じて、人同士のコミュニケーションがより促進される豊かな社会の実現を目指しています。「日常のタスクをAIに委ね、人が対面でのコミュニケーションに集中できる社会」を作ることが私の目標です。
- 専門領域: LLM Agents, Browser Agent, RAG, Edge AI, UI/UX Design.
- 拠点: 日本 | 言語: 日本語(母国語)/ 英語
「オーケストレーターでマルチエージェントを統合し、長期メモリと短期メモリを組み合わせることで、タスク成功率を1.7倍に改善しました」
卒業研究にて、5種の専門エージェント(RAG、Web操作、IoT、Scheduler、Orchestrator)を統合した「マルチエージェント・オーケストレーションシステム」を完遂。その研究は、NCSP'26(ハワイ・ホノルル開催)において発表しました。Web完結だけでなく、ハードウェアを含めた実世界へのフィードバックまでを一つのシステムとして統合・評価した経験は、国内でも稀有な実績です。
- Resume: https://github.com/kota-kawa/resume
- Portfolio: https://project-kk.com/
- LinkedIn: https://www.linkedin.com/in/kota-kawa/
- Email: kota7kawagoe@gmail.com
現在実際にアクセスして利用可能なサービスです。
- [よろずの窓口] (https://chat.project-kk.com/) — AIコンシェルジュ(マルチエージェント)。
- [ChatCore-AI] (https://chatcore-ai.com/) — フルスタックAIチャット基盤&プロンプト共有。
- [FS-QR] (https://fs-qr.net/) — ファイル共有 & QR。
-
Browser-Agent — 自然言語でWebを操作する
- 概要: FastAPIベースのWebインターフェースを持つブラウザ自動操作エージェント。自然言語の指示でタスクを実行し、noVNCでリアルタイム監視が可能。
- 性能: WebArena Shoppingタスク成功率: 32.6%。
- 使用技術: Python, FastAPI, TypeScript, Docker.
-
ChatCore-AI (Live) — 本番運用を見据えたAIチャット基盤
- 概要: メール認証、永続チャット、プロンプト共有を備えたフルスタックAIチャットシステム。GroqとGeminiを統合。
- 使用技術: Python (FastAPI), Next.js, PostgreSQL, Docker.
-
Symphony Agent Conductor — (自律型オーケストレーションプラットフォーム)
- 概要: LangGraphを用いて専門エージェント(ブラウザ操作、IoT制御、RAG)を統合する自律型オーケストレーションプラットフォーム。デジタルタスクと物理デバイス制御のシームレスな連携を実現。
- 使用技術: Python, Flask, LangGraph, browser-use, IoT (Jetson/Pi), OpenAI/Gemini.
- 慶應義塾大学大学院 湘南藤沢キャンパス(SFC)
- 政策・メディア研究科 サイバーインフォマティクス専攻 修士課程
- 神奈川県 | 2026年4月 – 現在
- 神奈川工科大学
- 情報学部 情報ネットワーク・コミュニケーション学科
- University of North Alabama — 交換留学プログラム修了
- アラバマ州, 米国 | 2023年5月 – 2024年5月
| カテゴリ | 技術 |
|---|---|
| プログラミング言語 | Python, TypeScript, SQL |
| Web / アプリケーション基盤 | FastAPI, React, Next.js, Tailwind CSS, PostgreSQL, Redis |
| AI / LLM エンジニアリング | RAG, Multi-Agent Systems, LangChain, LangGraph, browser-use |
| インフラ | Docker, AWS (EC2, VPC, Systems Manager), Linux, Nginx, Git, GitHub Actions, pytest |
| IoT / ハードウェア | NVIDIA Jetson Orin Nano, Raspberry Pi 4/Pico W |
- 開発: OpenClaw・Hermes Agentに着想を得た統合チャットインターフェースを開発。Ubuntu上でブラウザからOS操作まで自然言語で制御し、パーミッション管理と実行制御で安全性と実用性を両立したエージェント基盤を設計。
- 開発: 本番環境でのAIエージェントの安定稼働・信頼性向上に取り組んでいる。
Last updated: 2026-05-18



