Hi, I am Daiki Yamada, CTO of VoicePing.
VoicePing was selected for the final evaluation of the J-StarX Silicon Valley Extended Program, organized by 500 Global and JETRO, and participated in the on-site program in Silicon Valley from October 28th to November 22nd.
First, let me briefly explain the program content: It’s designed for startups from pre-seed to Series A stages who are considering global expansion. The program is divided into three phases. In Phase 1, participants spend two weeks learning about marketing strategies such as ICP (Ideal Customer Profile), go-to-market strategies, and effective pitching techniques. After each company delivers their pitch, 15 companies were selected to advance to Phase 2. In Phase 2, companies work with global mentors to design strategic plans for North American expansion over a two-week program. Phase 3 involves a month-long visit to Silicon Valley, aimed at establishing a foothold in North America based on the plans developed in Phase 2. While Phase 3 includes several organized programs, companies primarily operate independently to achieve their individual goals, such as fundraising and customer acquisition in North America.
I’d like to share some of the events we participated in and our visits to several famous company headquarters during this acceleration program. Also, our CEO Akinori Nakajima has written a blog about Phase 1, which you can check out if you’re interested. https://medium.com/@atyenoria/insights-from-500-globals-j-starx-program-on-nsm-pmf-and-pitching-to-vcs-fd332776f63b
First, a brief self-introduction: I’ve been working at VoicePing as a full-stack engineer for over a year and a half, involved in developing the web application’s frontend and backend, creating Speech-To-Text AI models, and other general features. We currently have 8 employees, all of whom are capable of coding – a rather unique environment where we can grow together through friendly competition.
Currently, VoicePing is recruiting AI Engineers and Web Development Engineers. If you’re interested in working with cutting-edge language models, developing high-performance speech recognition and translation models, or designing, developing, and operating complex, high-performance web applications, please feel free to contact us.
October 28-30: We attended TechCrunch Disrupt 2024 in San Francisco, a major tech event with over 10,000 participants (about 50% startups) and approximately 1,200 investors, focused on discovering new innovations. I’ve written a separate blog post about TechCrunch. https://voice-ping.com/blog/en-blog-techcrunch-disrupt/
Meta HQ
On November 4th, thanks to our employee’s friend who works as an engineer at Meta, we received a tour of the Meta office. He has been working as a Software engineer for 5 years. His main role is building the OS for Oculus Quest. There should be a lot of optimizations in order to make a stable OS.
I was amazed when I first entered Meta headquarters because it wasn’t just an office – it was an entire town. It had everything you need for daily life: numerous restaurants and cafes, a free bicycle repair shop, a barbershop, a sports center, and more. The restaurants featured diverse cuisines including Indian food, ramen, hamburgers, and ice cream, reflecting various cultures. While restaurants are mainly open during lunch, some operate during breakfast and dinner hours as well. I heard that a former Disney designer was responsible for the office design.
During our visit, we had BBQ lunch and Sprite. The American-style meat on the bone was delicious.
Over lunch, we learned about Meta’s development system. What particularly interested me was the role of UX researchers, who constantly investigate new UX trends online. They compile their findings into reports, which then inform new app development. Meta manages various popular social networks like Instagram and WhatsApp, and it was clear they put special emphasis on UX and performance metrics like speed. While Meta has many employees, each team typically consists of about 12 people, focusing on quality over quantity.
Meta headquarters even has a small arcade with about 10 arcade games including Fist of the North Star and Tetris. It was surprising to see these retro Japanese elements. We learned that employees purchased and installed these games themselves.
There was also an interesting vending machine – but not an ordinary one. It dispensed free work-essential electronics like headphones, USB cables, and keyboards. No need to spend your own money on these items anymore.
The Meta office campus is so huge it’s like a small city – you need to either drive your own car, borrow a bicycle, or take shuttle buses to get around. The scale is so different from Japan that it’s hard to find the right words to describe it. This is the shuttle bus stop area. Similar to a train station, there are screens displaying the bus arrival times.
Some facilities have rooftop walking paths. These are perfect for refreshing when you need a break from office work. I heard that managers and their team members often have walking meetings instead of sitting in an office.
The Meta office interior features stacked cube-shaped rooms arranged in an offset pattern, creating a stylish design. This probably contributes to employee motivation.
The overall Meta campus consists of about 60 buildings. While it might be hard to grasp the scale from the map alone, imagine several university campuses combined into one. That’s the kind of scale we’re talking about.
There is some anecdotal background story about not following the legacy company that won’t take risks and create something innovative like Sun.
Japan Startup Program Open House
VoicePing exhibited at the Japan Startup Program Open House by J-StarX, held at the SVB Experience Center from 18:00 to 21:00 on November 4th. A total of 15 startups exhibited, representing various sectors including AgriTech, SaaS, and BioTech. The event attracted 150-200 attendees, including investors, entrepreneurs, and employees from prominent companies like Google.
Our CEO Nakajima delivered a 30-second pitch. Attendees showed great interest in VoicePing’s real-time voice translation service.
Despite the brief three-hour duration, we were able to demonstrate VoicePing and answer many questions from visitors. People were particularly impressed by VoicePing’s ability to display transcriptions and translations in real-time. As the development team put considerable effort into optimizing speed, it was gratifying to see users recognize this achievement. The most frequently asked question was why VoicePing’s AI model is particularly strong with Asian languages. There are two main reasons: First, VoicePing’s development team includes native speakers of Japanese, Chinese, Korean, and other Asian languages. It’s impossible to improve accuracy without truly understanding the language. Second, we possess a large volume of meaningful business meeting data. High-quality data is crucial for AI training, and for business-focused AI, it’s particularly important to use data from actual business contexts.
VoicePing is an all-in-one tool that provides real-time transcription and translation not just for web conferences, but also for face-to-face meetings, global events, and even in offline environments without network connectivity. VoicePing breaks down language barriers in any environment, supporting our customers’ international business expansion.
Moon Creative Lab
November 6: We visited Moon Creative Lab, a shared office space operated by Mitsui & Co.’s startup program. We met Hoan, a senior product designer in Palo Alto. We worked with him through the Moon Boost program https://www.mooncreativelab.com/boost . We talked about our product growth and consulted about the challenges we are facing for a simple and concise user experience to target the one Ideal Customer Profile. We also have got to know how they are going with local Entrepreneurs. Local networking event from them https://www.eventbrite.com/e/networking-happy-hour-at-moon-creative-lab-tickets-1076026768029 will foster the SF ecosystem.
Google HQ
November 7: We had an hour-long discussion at the Google Visitor Center with a Product Manager who had worked on Google Translate. We received feedback about VoicePing’s current status and strategies for expanding into North American and other markets. Also we could be heard the secret source why Google is always No.1.
The reason why Google is overwhelming among competitors is from brand recognition and product quality. When Google Translate emerged, there was no other product than it in terms of accuracy and user-friendly user experience. Having the automatic language detection to translate was a killer feature.
The secret source of having too many billing users app is the company culture. Doing a hackathon and killing 97 and taking only 3 out of 100. Embedding the culture of creating something new always regardless of the company size is what we really have to follow.
Google’s philosophy “If no reason to make it a commercial app, provide it with a free” is user-centric. We hopefully provide the usable app for free as long as we can sustain it.
November 12: We visited Notion and Stripe headquarters.
Notion HQ
We are a heavy user of Notion. It’s very honor to visit the Notion office. At Notion, we learned about their global marketing strategies and Value Position. They emphasized the importance of balancing product refinement based on customer feedback with innovative feature development, while focusing on perfecting key features rather than trying to implement everything.
Stripe HQ
Our business system is based on the Stripe Platform. It’s very exciting to visit the Notion office. At Stripe’s headquarters, we had lunch. It was kebab day, with a buffet-style system where employees can choose their portions. The cafeteria was packed during lunch hours, demonstrating the impressive scale of their workforce.
Nvidia HQ
November 13: We visited Nvidia’s headquarters. Our product is based on so many Nvidia GPUs. We were really looking forward to seeing the Nvidia Culture. Fortunately, we got into the Nvidia inception program https://www.nvidia.com/en-us/startups/ That will lead to collaboration with Nvidia. As a AI startup, it’s very horning to have this opportunity. There were features an elegant design with intentional triangular patterns throughout for ventilation. The office windows automatically adjust based on temperature, showing their commitment to creating an optimal work environment.
PayPal HQ
November 14: We visited PayPal’s office and had a discussion with a former Airbnb lawyer now working at PayPal. The most memorable topic was the cultural differences between San Francisco and other regions, particularly Japan. While Japanese startups often focus on efficiency, San Francisco startups tend to pursue innovation. This mindset extends to investors and venture capitalists. Job mobility is also different, with average tenures of two years or less being common in San Francisco. Companies respond by offering perks like free meals and comfortable offices to retain talent.
ByteDance
We visited the ByteDance office in San Jose to attend the ByteDance Open Source Meetup held on November 21st. There was a presentation about Monolake, a framework for building high-performance systems by leveraging Rust’s zero-cost abstractions. Monolake was developed by ByteDance and can be used to create high-performance proxies and gateways, with results showing better performance than Nginx. Monolake is open source and available on Github, so if you’re interested, you can check it out at https://github.com/cloudwego/monolake.
There was also a discussion about Whisper, an automatic speech recognition model developed by OpenAI, covering what Whisper is, how it can be utilized, and optimal approaches for fine-tuning. Since VoicePing also creates high-performance speech recognition models by fine-tuning Whisper, we were able to gain very valuable information.
Yosemite National Park
I visited Yosemite National Park on a holiday. While you might imagine a “park,” it’s incomparably larger than any national park in Japan. You basically need a car to move around within the park, and the scale of natural features – the rocks, mountains, and trees – is many times larger than what you see in Japan. I arrived at Yosemite right around sunset, so the mountains were captured beautifully in the light.
Last
Although I’m not particularly good at English, I was able to keep up with conversations by using VoicePing’s mobile app to transcribe and translate English conversations in real-time. Additionally, after meetings, it automatically creates summaries, allowing me to quickly review what was discussed. You can use it for free for 60 minutes, so please try it out or visit VoicePing’s website if you’d like to learn more!
Summary
This was my first time visiting US, and I was shocked by various aspects. For example, the sense of scale was surprising. Places that looked close on Google Maps could actually take hours to reach. Another aspect was the culture of valuing personal connections. At events like TechCrunch and others, they always allocated plenty of time for networking, and it was common practice to maintain these connections through LinkedIn. In Silicon Valley, there were frequent events attended by investors and entrepreneurs happening all over the place, with everyone actively working towards creating the next unicorn company. This made me realize that this might be the key difference between Japan and US in terms of why unicorn companies emerge or don’t emerge. I also had the opportunity to visit the headquarters of many famous tech companies, which gave me a real sense of their scale and provided valuable learning opportunities about various aspects like their marketing strategies.
Currently, VoicePing is recruiting AI Engineers and Web Development Engineers. If you’re interested in working with cutting-edge language models, developing high-performance speech recognition and translation models, or designing, developing, and operating complex, high-performance web applications, please feel free to contact us.
