​Topic

​Tokyo AI (TAI) presents a session on "AI x Agents", an ever-green for 2025, with four AI-driven startups talking about their applications.

​Talk 1: Supercharging Your Startup with AI Agents as Pair Programming Partners

​Speaker: Eduardo Gonzalez, CEO & Co-founder, Xpress AI

​In the startup world, speed is critical—especially when it comes to building and shipping quality code. Eduardo will share their innovative approach to leveraging AI agents as pair programming partners in a methodology they call Omega Programming, inspired by Extreme Programming principles.

​Rather than struggling with LLM-generated code completions that are often subtly incorrect, Omega Programming reframes the relationship. You treat the AI as a coding partner, actively leading and mentoring it in the pair programming style, guiding the agent in the right direction.

​By combining test-driven development with a focus on high-level architecture, this approach can significantly boost productivity, allowing developers to spend less time fixing errors and more time creating impactful solutions.

​Talk 2: Accelerate Your AI Agents: Seamless Internal Agent Deployment and Sharing

​Speaker: Naoto Shibata, Co-founder & CEO, Morph.

Morph is an open-source Python framework designed to streamline the creation and deployment of internal tools—including AI agents—without requiring complex infrastructure setup.

​In this session, we’ll demonstrate how you can rapidly develop AI-driven applications and securely share them with your team using just a single command. Additionally, we’ll explore Morph’s built-in role-based access control, which makes managing user permissions effortless, ensuring that the right people have access to the right features.

​Discover how Morph accelerates the delivery of AI agents, enhances collaboration, and ensures robust security within your organization.

​Talk 3: Real Messy Imperfect World End-To-End Agent

​Speaker: Sam Yu, Co-founder and CTO, YuzuAI.

​While AI agents show promising capabilities in controlled environments, deploying them in real-world applications presents numerous unforeseen challenges. Through practical implementation, we'll uncover critical obstacles including non-uniform data source formats, content access restrictions like paywalls, inconsistent AI model behaviors across different tasks, and the need to understand and compensate for specific LLM strengths and limitations.

​Sam will discuss architectural decisions that enhanced system resilience, methods for handling unexpected edge cases, and strategies for maintaining reliable operation across multiple service integrations.

​Key insights will include understanding fundamental LLM limitations and designing architectures around them, implementing systems that augment rather than override LLM capabilities, and the importance of measured, incremental automation rather than attempting comprehensive solutions immediately.

​Talk 4: Show, Don’t Tell: How Interfaces Shape the Future of AI agents

​Speaker: Kenn Ejima, Founder & CTO, Admit AI.

​A week ago, nothing was more low-status than being a “GPT wrapper” startup. DeepSeek just changed that perception. Frontier models are becoming commodities, but user habits are not.

​With frontier models rapidly becoming commodities, the real competitive edge lies in creating intuitive, user-first interfaces that redefine how people interact with AI agents. Drawing on the tech industry’s history and long-term trends, Kenn will share insights on designing interfaces that engage users, shape habits, and ensure widespread adoption of agentic apps.

​Join to discover why a great user experience is the ultimate moat and how it can elevate agentic apps from niche tools to mass-market solutions.

​​Our Community

​​​​Tokyo AI (TAI) is a community composed of people based in Tokyo and working with, studying, or investing in AI. We are engineers, product managers, entrepreneurs, academics, and investors intending to build a strong “AI coreˮ in Tokyo. Find more in our overview: https://bit.ly/tai_overview

​​Schedule

​18:30 Doors open

​19:00 - 20:30 Speaker sessions

​20:30 - 21:00 Networking

​21:00 Event ends

​Sponsor

​AWS is kindly hosting us for this session.