Short Bio
Ken Imoto is a WebRTC & Voice AI engineer and founder of Propel-Lab. He is the author of the LLMO Framework and has published 30+ Kindle editions across 4 languages (plus 13 Zenn Books) on AI development, context engineering, and AI search optimization.
Medium Bio
Ken Imoto is a WebRTC & Voice AI engineer building AI-native organizations at Propel-Lab. He is the creator of the LLMO (Large Language Model Optimization) Framework, which helps businesses optimize their content visibility in AI-powered search engines. Ken has published 30+ Kindle editions across English, Japanese, Portuguese, and Spanish — including "Practical Claude Code," "Harness Engineering," "Knowledge Graph Practical Guide," and "LLMO Practical Guide" — plus 13 Zenn Books. His research spans AI text analysis, vision-language model bias, and voice AI latency optimization, with 4 papers published on Zenodo. His technical writing on Qiita spans 120+ articles, and combined with Zenn his writing has attracted over 400,000 pageviews.
Full Bio
Ken Imoto is a WebRTC & Voice AI engineer and founder of Propel-Lab, where he builds AI-native organizations powered by LLMs, automation, and distributed agents.
He is the creator of the LLMO (Large Language Model Optimization) Framework, a systematic approach to optimizing content visibility in AI-powered search engines like ChatGPT and Perplexity. The framework is documented in his book "Why ChatGPT Ignores Your Website: The LLMO Practical Guide" and the companion site llmoframework.com.
Ken has published 30+ Kindle editions across English, Japanese, Portuguese, and Spanish (plus 13 Zenn Books) spanning AI development, context engineering, knowledge graphs, MCP security, and AI search optimization. Notable titles include "Practical Claude Code — Context Engineering for Modern Development," "Harness Engineering — From Using AI to Controlling AI," "Knowledge Graph Practical Guide — Structure Your Data, Sharpen Your AI," and "MCP Security in Practice."
His research includes four papers on Zenodo: "When Free Executors Cost More" (showing that the canonical strong-orchestrator + cheap-executor pattern is the most expensive cloud arm on every code-repair task because prompt-cached re-reads dominate execution savings), "Excess Vocabulary in Japanese AI-Generated Text" (analyzing 651 statistically significant excess words across 7 LLMs), "AI Text Slop" (measuring stylistic convergence in Japanese technical writing), and "AI Blue" (quantifying color recognition bias in vision-language models). The latter three are quantitative analyses linking specific LLM behaviors to the broader "AI Slop" phenomenon; the first is an empirical Pareto comparison of agentic coding configurations.
Ken is currently researching voice AI latency optimization (the 300ms barrier), context engineering for multi-agent AI systems, and generative agent simulation.
Featured Work
- "Running Qwen 35B on 12GB VRAM" — 60,000+ PV on Qiita (2026-06)
- "Managing Everything on GitHub in the AI Era" — 45,000+ PV / 241 LGTM on Qiita (2026-03)
- "Six-Stage Code Review Reveals the New Human-AI Division of Labor" — 284 Likes on Zenn (2026-05)
- "Harness Engineering: Five Companies, Five Definitions" — 130 Likes on Zenn (2026-04)
- Top posts on TabNews (Brazil PT-BR) — earning 10–12 tabcoins each (2026-05–06)
Key Statements
- "Engineers don't need a faster model. They need a harness." — Harness Engineering
- "AI search citation is decided by passages, not pages." — LLMO Practical Guide
- "An LLM doesn't get smarter — your structure does." — Knowledge Graph Practical Guide
- "What changes Claude Code's output isn't the model — it's the context you frame it with." — Practical Claude Code
- "The 300ms latency boundary is where 'AI conversation' becomes 'AI assistant.'" — Voice AI 300ms
- "Context Engineering turns LLMs from liars into experts." — Context Engineering in Practice
- "LLMO is the discipline of making content quotable by machines, not just findable." — LLMO Quickstart
Media Kit
Profile Photo
Download (JPG)Usage Guidelines
- Do not crop or alter the photo
- Use "Ken Imoto" for attribution
- Link to kenimoto.dev when possible
Side Projects
- 🥃 legacydram.com — Whisky curation read through an engineer's lens
Open Source
compact-ops — Claude Code plugin · MIT
A transparent plugin that keeps Claude Code sessions coherent across context compaction: structured state capture before /compact, recovery injection right after compaction and on --resume, plus a self-contained usage warning. Derived from u-ichi's compact-plus.
domain-pre-flight — Python CLI · MIT
A pre-flight check before you register a domain for a new site or app. Structure, history, typosquat, multi-language semantics, LLMO, and trademark deeplinks in one command.
voice-clone — Python · MIT
A voice-cloning tool built on Qwen3-TTS. Clone a voice from ~3 seconds of audio and have any text read in that voice. WSLg recording, multi-language.
speech-habit-lens — Python CLI · MIT
A 1-minute speech habit analyzer that combines AmiVoice ESAS (20 acoustic emotion parameters) with an LLM. Three layers — acoustic, textual, and the cross-layer that links body and language — rendered as a Markdown report.
persona-hub — TypeScript + Python · Apache-2.0
A two-part persona hub: a lightweight TypeScript SDK that scores quiz answers locally, and an optional FastAPI service that persists results so the same user can carry their profile across services. Architecture borrowed from LaunchDarkly + Segment Engage, applied to persona aggregation. Pre-alpha.