AIEEV at TechCrunch Disrupt 2026: Notes from Startup Battlefield 200
Hello from AIEEV 👋
This week we traded the office for San Francisco. AIEEV was selected for the Startup Battlefield 200 at TechCrunch Disrupt 2026, and we're here introducing Air Cloud to people seeing it for the first time and hearing what kinds of AI products they're building.
Most of our posts focus on product and features. This one is about the team that brought that product to the show: a packed schedule that started before the main event, getting our pitch ready, and the people we met at our booth.
A busy week before the event even started 🥹
Our first stop in San Francisco was a training and mock pitch session hosted by Shinhan Square Bridge. We heard about the local market and went over our pitch with mentors. Looking at our business through the eyes of US customers and investors, and getting direct feedback on our delivery, helped a great deal as we prepared for the main event.
It was also great to meet other Korean startups taking on the US market. Introducing our products to each other often turned into conversations about how to reach customers here. We work in different fields but share many of the same challenges. Talk about how each team was preparing and which customers we hoped to meet went on for quite a while.
Between networking and adjusting to the time difference, we kept practicing the pitch. We had two minutes to introduce AIEEV and Air Cloud. Explaining our business to a US audience in that little time meant hard choices about what to keep and what to cut. We worked in the mentors' feedback, rehearsed out loud, and kept refining the pitch right up to the event so the customer problem and how Air Cloud solves it would come across clearly.
Introducing AIEEV in two minutes
On October 14, AIEEV Co-Founder Karen Kim took the stage at the Startup Battlefield 200 Pitch Showcase. In front of startups and investors from around the world, she spoke about the infrastructure costs AI companies face and how AIEEV addresses them by connecting underused GPUs into a cloud.
Air Cloud grew out of two very different situations. Teams building AI services need GPUs at every step: to run models, to test products, and to keep up as their user base grows. That inevitably turns into a real cost burden. Meanwhile, corporate research labs and edge data centers often have GPUs that sit underused. Where GPUs are needed they're expensive, and where they already exist they go unused. AIEEV connects these GPUs and offers them as a cloud. Finding and connecting GPUs is only part of the job, though. Customers need to deploy their own models, send requests, and get results back before any of it is useful in a real service. What we build is the environment that makes scattered GPUs usable in exactly that way.
In the pitch, we covered how Air Cloud lowers infrastructure costs for AI teams and how companies that own GPUs can put idle capacity to work. Karen closed by inviting both groups to our booth to talk through how Air Cloud could fit their own environments.
What AI teams told us at the booth
Visitors to the AIEEV booth at E42 were building all kinds of services, from construction AI and game graphics generation to voice AI and braille accessibility technology. After hearing what they were working on, we invited them to try the Air Cloud demo themselves. As visitors picked a workload and a GPU, we walked through the deployment flow and estimated costs alongside them.
The demo made our conversations more concrete. Teams running on cloud credits were already thinking about what they'd pay once those credits ran out. With their current setup next to the demo's cost comparison screen, we talked about whether Air Cloud could be an alternative. People preparing for hackathons or side projects asked whether they could start small with only what they need. Once we heard what they wanted to build, we could point them to the right GPU and explain how to get started.
Some questions went beyond what a screen can answer. With people from the public sector, we talked in more detail about where data is processed and how the environment is operated. Because every visitor tried the demo and asked their own questions, the same demo led to a different conversation with each team. For us, it was a chance to go beyond presenting features and hear which factors actually matter when teams choose a service.
We connected on LinkedIn with everyone who tried the demo and handed each of them one of the keycap keychains we made for the event. We hope these booth visits are only the beginning and that we'll keep in touch long after the event.
One more day to go
As we write this, the final day of the event is still ahead. While preparing for the training and the pitch, we were focused on what we wanted to say. Running the booth, we found ourselves doing far more listening. Thank you to everyone who told us where and how they'd like to use Air Cloud.
We'll be at E42 again tomorrow. If you can't make it in person but are curious about Air Cloud, feel free to reach out 🤗
