フリーランスエンジニア(東京)。AI/LLM とクラウド自動化を中心に、実用プロダクトを作っています。
I build AI/LLM products and cloud automation, with a focus on reliability, fairness, and safety.
- Portfolio(実績・スキル・お問い合わせ): https://naoya-yasuda.github.io/portfolio2/
- Another Star LLC(GENIAC-PRIZEのために設立した合同会社): https://www.another-star.jp/
AIエージェントの 安全性・評価・検知 をテーマに、プロトタイプ開発と検証を進めています。
- https://github.com/Naoya-Yasuda/secure-ai-agent-matching-platform
AIエージェント同士をセキュアにマッチング/仲介する基盤(エージェントストア側を担当) - https://github.com/Naoya-Yasuda/browser-agent-detector
顧客クラスタリング × ブラウザ操作情報で “AIブラウザ操作” を検知する多層防御(初期クラスタリング設計・実装を担当)
- Cyber University(サイバー大学)を 社会人学生として卒業(AIテクノロジーコース)
- マナビDX 地域協業プログラム 参加
- GENIAC-PRIZE 参加(AIエージェント/評価・安全性など)
- https://github.com/Naoya-Yasuda/corporate_bias_study
サイバー大学の卒業研究:AI検索サービスにおける企業優遇バイアス - https://github.com/Naoya-Yasuda/HackU2023_project
Hack U 2023 ONLINE 最優秀賞・Happy Hacking賞(盲導犬アプリ) - https://github.com/Naoya-Yasuda/MANABI_DX_QUEST_2023_PrePoC
マナビDXクエスト2023(地域協業プログラム)データ分析
- LLM / RAG / AI Agent Safety
- Evaluation & Metrics(信頼性・公平性・再現性の評価)
- Data / Analytics(GA4 / Snowflake など)
- Automation(n8n など)
- Collaboration / Freelance or contract work(フルリモート可)
- PoC開発、プロトタイプ、評価設計、クラウド基盤整備 など
- X(DM): https://twitter.com/yasutyNhoffman/
- GitHub: https://github.com/Naoya-Yasuda
- Zenn: https://zenn.dev/family_chicken
- Qiita: https://qiita.com/2301330039zz
- Hackathon winner: https://hacku.yahoo.co.jp/hacku2023_online/index.html
- Research (corporate bias in AI search): https://corporate-bias-study.streamlit.app/


