ChatGPT DeepReseach (2025/02/22 20:00) ChatGPT DeepReseach (2025/05/24 20:00)
Tech Evangelist · Hardware/Software Engineer · Founder
X (Twitter): @GOROman
Areas: AI/LLM · Edge-AI Devices · XR · DIY Drones · Community Building
| Year | Role | Org / Product | Brief |
|---|---|---|---|
| 2025– | Manus Fellow (Japan) | Manus | Evangelizing the “AI agent” paradigm and running hands-on community events. |
| 2010– | Founder & CEO | XVI Inc. | Building XR and AI prototypes; open-sourcing tools for the global dev community. |
| 2023–2025 | Director & CTO | ViXion Inc. | Supporting R&D and go-to-market for the autofocusing smart-glasses “ViXion01S.” ViXion Blink & OpenBlink |
| 2014-16 | Partner Engineering Specialist | Meta (Facebook) | Launched Oculus,VR products in Japan and built dev relations. |
| Project | Year | Stack / Tech | My Role & Impact |
|---|---|---|---|
| KANTAN Play | 2024 | ESP32 · Godot Engine | Co-created a one-button music gadget; built the web app prototype and advised on firmware/UX. Successfully crowdfunded. |
| StampFly | 2023-25 | M5Stamp S3 · C++ · LLM-assisted design | Core dev of an open-source palm-size drone. Leveraged ChatGPT for architecture docs, firmware scaffolding, and PID tuning. |
| LLM Cardputer | 2024 | M5Stack Cardputer · Module-LLM | Turned a card-sized SBC into a local ChatGPT client; showcased fully offline LLM chat on a pocket keyboard. |
| SubtractProxy | 2024 | Node.js · Ollama · OpenAI API | Wrote a MIT-licensed AI-filtering HTTP proxy that rewrites/blocks web content via local LLMs. |
| HAH Dev-Flow | 2024 | GitHub Actions · OpenAI | Demoed the “Human–AI–Human” workflow: PR templates, code review, and task routing auto-generated by ChatGPT. |
- ScanSnap User Meetup 2025 – Keynote on “LLMs and Personal Knowledge Work.”
- Regular Twitter Spaces / Discord AMAs on Edge-AI, IoT × AI, and Dev Productivity.
- Contributor & mentor in the M5Stack, XR Consortium, and indie-dev communities.
Python · C/C++ · Ruby · Node.js · ESP-IDF · LLM orchestration
Generative AI, Edge inference, XR hardware, rapid prototyping, community-first product design.
I believe in a world where “one person → one AI agent” becomes reality.
My mission is to blend hardware, software, and large-language models so that anyone—regardless of skill level—can create, explore, and play with new ideas at the speed of thought.
Let’s build the future together—pull requests welcome!