The Story So Far

Current Chapter

Starting a new phase at Capgemini to understand enterprise-scale AI problems. Maintaining projects like JimakuAI on the side - because working tech that helps people deserves to live. Building towards financial independence through deep technical expertise and enterprise knowledge.

The Startup Chapter

Founded JimakuAI for real-time translation between Japanese and English. Built robust technology with sub-second latency and high throughput, now being used in church sermons. The tech worked, but I learned hard lessons about market fit: people with burning hair problems will pay even for imperfect solutions. Key insight: Pre-sell before building, and know your customers deeply - they're the ones who will pay even for a basic version.

The AI Engineer Chapter

Led AI engineering at AI Inside, building a team from ground up and scaling from research to production. Created a NO-CODE AI platform that democratized AI for non-technical users. Specialized in optimizing AI systems through rigorous benchmarking - analyzing model performance, GPU utilization, and cost-effectiveness across different deployment scenarios. This work led to 10-100X improvements in cost-performance ratios.

Developed deep expertise in AI infrastructure optimization: built comprehensive benchmarking suites comparing cloud APIs vs self-hosted models, analyzing tradeoffs between latency, throughput, and cost. Implemented MLOps practices with Kubeflow pipelines, enabling rapid experimentation while maintaining production reliability. Key achievement: Creating frameworks that helped teams make data-driven decisions about AI deployment strategies.

The Academic Chapter

PhD in Mathematics from Nagoya University, specializing in harmonic analysis and nonlinear PDEs. The deep technical foundation helps me approach AI problems systematically. The academic journey taught me to love solving hard problems, but I realized I wanted to build things people actually use.

Why I Build

My goal is maximizing positive impact through technical leverage. The path: Learn enterprise AI needs, build connections with decision-makers, and identify gaps where deep technical solutions can create real value. Following Naval and Munger's principles about leverage and incentives, while maintaining focus on problems worth solving.

Connect

Looking to connect with builders who care about real impact, not just the next funding round. Whether you're working on enterprise AI transformation or building tools that matter, let's talk. Especially interested in problems where deep technical expertise meets practical business needs.