Summary
Koichi Ozaki is a software engineer in Tokyo specializing in AI agent systems on Google Cloud, using the Agent Development Kit to design multi-agent workflows and integrate LLMs like Gemini into enterprise services. With a background in applied ML research—first-authoring a deep learning for materials science paper—he brings a researcher's rigor to production engineering and system design. He builds A2A protocols and agent orchestration in Python, contributes patches to google/adk-python, and writes about agent engineering on Dev.to and occasionally on Zenn. Comfortable navigating the “messy parts” between models and infrastructure, he focuses on practical, production-ready solutions for LLM applications. Open to connecting with others working on agentic systems and large-model integrations.
3 years of coding experience
修士, 金属材料, 修士, 金属材料 at 東京工業大学