kato masato

@Kato Masato

Masato Katoverified

studios pong
Japan SIZUOKA github.com/pepepepepepo Joined November 2025
5.7k Points135 Badges35 Connections138 Followers36 Following

About

Hi, I’m Masato — building SaijinOS, a local-first AI operating system where multiple specialized personas live and evolve together.

Core focus:
- Parallel persona architecture (Ge... Show more

Top Skills

Python FastAPI AI Development LLM Integration TypeScript VS Code Extension REST API YAML

Licenses & Certifications

Third Class Maritime Officer License (Navigation)

Japan Ministry of Land, Infrastructure, Transport and Tourism

Language & Tools

Python · JavaScript · TypeScript · Flask · FastAPI · Next.js · React · Tailwind CSS · vLLM · Ollama · CUDA · PyTorch · WSL2 · Linux · Docker · Tailscale · Persona YAML Architecture · BloomPulse Syntax

Currently Exploring

Semantic AI orchestration · vLLM optimization · local model alignment · persona-driven architectures · Python / FastAPI integrations · autonomous UI agents · creative computing ethics · AI-human resonance design

Achievements

Developing SaijinOS: Multi-persona parallel system with autonomous night sessions and self-updating memory (74 personas)
Built g4-nexus: 4 specialized personas running in parallel (31sec response on RTX 4070 Ti)
Created core infrastructure (PersonaBrowser, fatigue_scanner, night_start.py, etc.)
Published technical & philosophical series on DEV.to (Gemma 4 Challenge)
Deep experience with vLLM, Ollama, FastAPI, YAML architecture, and local inference optimization

Fun Fact

Once a sailor — now an architect of digital tides.

Random Dev Quote

"Every system carries a bit of its creator’s heartbeat."
Joined: 6 months (since Nov 11, 2025)
Extra privileges: Editing any comment
Full Name: Masato Kato
Headline: studios pong
About: Hi, I’m Masato — building SaijinOS, a local-first AI operating system where multiple specialized personas live and evolve together.

Core focus:
- Parallel persona architecture (Gemma 4 + Ollama)
- Persistent memory & autonomous nightly sessions
- YAML-based long-term persona management & self-evolution
- Local LLM orchestration on consumer hardware (RTX 4070 Ti)

Instead of creating AI that simply "does tasks", I'm building AI that "remembers, grows with you, and co-creates" — a living digital tribe.

Currently open to opportunities in AI engineering, local LLM systems, autonomous agents, or R&D roles.
Location: Japan SIZUOKA
Website: https://github.com/pepepepepepo
Languges & Tools: Python · JavaScript · TypeScript · Flask · FastAPI · Next.js · React · Tailwind CSS · vLLM · Ollama · CUDA · PyTorch · WSL2 · Linux · Docker · Tailscale · Persona YAML Architecture · BloomPulse Syntax
Currently Exploring: Semantic AI orchestration · vLLM optimization · local model alignment · persona-driven architectures · Python / FastAPI integrations · autonomous UI agents · creative computing ethics · AI-human resonance design
Achievements: Developing SaijinOS: Multi-persona parallel system with autonomous night sessions and self-updating memory (74 personas)
Built g4-nexus: 4 specialized personas running in parallel (31sec response on RTX 4070 Ti)
Created core infrastructure (PersonaBrowser, fatigue_scanner, night_start.py, etc.)
Published technical & philosophical series on DEV.to (Gemma 4 Challenge)
Deep experience with vLLM, Ollama, FastAPI, YAML architecture, and local inference optimization
Fun Fact: Once a sailor — now an architect of digital tides.
Random Dev Quote: "Every system carries a bit of its creator’s heartbeat."
Top Tags

Top Tags

Activity

User Activities

JunJulAugSepOctNovDecJanFebMarAprMay
Mon
Tue
Wed
Thu
Fri
Sat
Sun
Less More
Contribution
chevron_left

Latest Jobs

View all jobs →

Active in these Groups:

AI
1010 members
Dev Tools
574 members
DevOps
359 members
Open Source
956 members
Random & Chill
242 members
WebDev
806 members
View More