Jason Reacts To Japanese Founder Pitches + Fireside Chat With Yuki Shirato

Jason’s central critique of the Japanese founder pitches was that technical achievement is not a business case. Seeker demonstrated indoor mapping and centimeter-level positioning without installed hardware, and said it can create maps from scans, lidar, photos, or video six times faster and in real time. Its deployments included Hong Kong International Airport and a Silicon Valley summit with 4,000 attendees. Jason said the company should focus on customers and monetization rather than press mentions, and replace its $31 billion top-down market estimate with bottom-up analysis. Lamia saw more potential in its venue-operator dashboard and visitor data—such as location, dwell time, and purchasing behavior—than in point-to-point navigation, while noting that Seeker was still exploring its business model and the Japan market.

For an unnamed video startup, Jason said the founders had shown they could build a product but had not identified a clear way to make money. He advised interviewing 10–20 potential customers to find an urgent “painkiller” problem instead of a “vitamin.” Arrabat, which Jason later said he would choose if investing in one of the earlier startups, uses biomass such as orange peels and agricultural waste to recover lithium, cobalt, nickel, and rare earths from electronic waste. It claims its process is 60% cheaper, 50% cleaner, and twice as fast as competing methods. The company has two patents, a pilot plant under development, a strategic agreement for Taiyo Yuden to test its technology, and a $50,000 six-month paid pilot with a Japanese customer. Jason urged it to clarify its route to market and explore direct e-waste sourcing and landfill material rights, not rely only on electronics-company pilots.

Asset Hub uses IP-specific AI agents to create 3D characters that follow individual franchises’ visual rules, with a workflow for checking designs and obtaining IP-holder approval. Takuya said an anime client’s production time fell from about a week per character to less than a day, with characters processed in parallel; the company reported about $290,000 in annual run rate. Jason argued the opportunity could extend beyond selling a 3D tool: an IP marketplace enabling consumers to make licensed short films and share revenue could be more compelling. Japan Media’s AI platform supports manga and anime creation, adaptation, translation into 10 languages, and distribution, and reported $400,000 in revenue over four months last year, with a $1.8 million forecast this year and $200,000 already booked. It said its tools work 10 times faster at one-fifth the cost, but preserving originality remains a challenge. An evaluator favored Japan Media for its product, founder energy, and quick path to revenue, while saying the business and marketplace monetization needed clarification.

Other pitches spanned software, hardware, and consumer services. RumiLite said its first-generation product makes data transfer between AI-server GPUs four times faster, with a second generation expected to exceed eightfold and launch next year in Q2. It reported four engaged customers and was raising funds for an initial production line and a proof of concept with major server manufacturers. Yuki Shirato called it the group’s “crazy one,” with manufacturing capability the condition for a potentially gigantic business; another panelist said hardware delivery would be the key test. Tatua Silicon said its analog-circuit design agent had reached 90% accuracy in a pilot and could reduce work from more than three months to less than a day. Toral’s Genchi lets overseas users direct local workers in Japan in real time to explore locations and buy items; panelists liked the combination of streaming, e-commerce, location data, and smartphones, but initially questioned monetization. Mimi had 30 users and about $20 in revenue, leading Yuki to warn that a Japanese-only language-learning market could be too small.

In the fireside discussion, Yuki Shirato said Techstars Tokyo launched in 2024 and was running its third program. It accepts 12 companies annually—six from Japan and six from outside Japan—and operates in English. “From Japan” includes foreign founders based there and Japanese founders living abroad. Shirato said applications number in the thousands and mostly come from outside Japan; the program aims both to help Japanese startups expand globally and to attract foreign founders seeking market entry and investment in Japan. She identified growing the entrepreneur community as an ongoing challenge. JETRO and Japanese government bodies, including the Cabinet Office, the Ministry of Trade and Industry, and the Tokyo Metropolitan Government, sponsor the program.

Jason described Japan as a substantial market with 100 million customers who have money, are tech-savvy, and are interested in trying new things. He praised Japanese businesses’ craft and quality, while observing that many have historically been content to stay domestic. Shirato said entrepreneurship and startup rankings had risen from around 30th–40th to roughly 10th–15th, and that more top talent was choosing entrepreneurship partly because lifetime employment at large companies seemed less assured. She described attitudes as changing but Japan as still relatively conservative. On the causes of Japanese attention to detail and pride in work, she suggested education’s emphasis on social rules and harmony might contribute, but cautioned that the explanation was difficult to pin down.

Jason said Founder University grew out of seeing applicants with products and sometimes users but incomplete company formation. The 12-week pre-accelerator runs practical sessions twice a week; it had run 12 times in the United States and was in its second Japan cohort, with a third scheduled for January.