Sakasegawa's Homepage

Latest Posts

• speech.cpp 0.8.4 released: quantized versions of every model, and more 2026-10-12
• How far can the OpenAI Decisions API go on image recognition tasks? A comparison with other methods 2026-10-11
• Building speech.cpp, a Speech Synthesis and Recognition Engine for Voice Conversation 2026-10-09
• I Built a Voice Assistant for the Desktop 2026-09-24
• Evaluating Antigravity Gemini 3.5 Flash and Cursor Composer 2.5 on HarnessBench 2026-05-24

→ See all posts

Projects

Computer Agent / RPA Agent

Agents built on top of computer-use

Article Computer Agent Demo RPA Agent Demo

Subtitle / Dubbing System

Multilingual video subtitling and AI dubbing system

Subtitle Demo Dubbing Demo (ES→JA) Dubbing Demo (DE→KO)

Virtual Try-On

Virtual try-on system

Demo

Hiragana ASR

Speech recognition model that transcribes Japanese speech as hiragana

Article Demo GitHub Model

spark-banana

A fast UI iteration tool combining Codex Spark and nanobanana

Article GitHub

notion-cli

Notion CLI tool for coding agents

Article GitHub npm

Featured Posts

Comparing 21 Japanese Handwriting OCR Models

2026-03-17

Harness Engineering Best Practices

2026-03-09

Skill Design Lessons from skill-creator, and Building an Orchestration Skill

2026-03-04

Coding Agent Roundup (January 2026)

2026-01-26

I Built 10 AI Apps and 3 AI Agents Solo

2025-08-07