soy

Patent lawyer turned AI engineer. Processed 4M patents with local LLM on RTX 5090. Building PatentLLM — AI-powered patent search. Also ranked #1 on Floodgate (shogi AI). Writing about local LLM etc.

Japan Joined on Mar 8, 2026 q07025a@gmail.com https://media.patentllm.org/

soy

Mar 8

I Built a Free Patent Search Engine with 3.5M US Patents — No Login, Powered by SQLite FTS5

#showdev #sqlite #python #llm

3 min read

Want to connect with soy?

Create an account to connect with soy. You can also sign in below to proceed if you already have an account.

Create Account

Already have an account? Sign in

soy

Mar 8

Operational Techniques for Automatically Starting vLLM, Flask, and cron with systemd Services in WSL2

#devtools #python #productivity

3 min read

soy

Mar 8

Achieving Bidirectional Integration of Streamlit Backend Flutter Frontend in a WSL2 Environment

#webdev #devops #infrastructure

2 min read

soy

Mar 8

A Regulatory Analysis Dashboard for Fast Searching NITE CHRIP Data using FTS5

#devtools #python #productivity

2 min read

soy

Mar 8

Searching Case Law PDFs with RAG — A Legal AI Search System using Gemini + SQLite FTS5

#devtools #python #productivity

3 min read

soy

Mar 8

google-generativeai google-genai Migration Guide

#devtools #python #productivity

2 min read

soy

Mar 8

Gemini 2.5 Flash x Nemotron 9B — Optimal Division of Roles for Cloud LLM and Local LLM

#ai #machinelearning #llm

3 min read

soy

Mar 8

Reduce API Costs for Large-Scale Document Analysis with Gemini Context Caching

#ai #machinelearning #llm

2 min read

soy

Mar 8

Skit: The Man Obsessed with Claude Code

#devtools #python #productivity

3 min read

soy

Mar 8

Building a Free Research Agent with DuckDuckGo Search + Local LLM

#ai #machinelearning #llm

2 min read

soy

Mar 8

A Daily Report System to Automatically Aggregate Claude Code + Gemini CLI Usage History Every Morning with Cron

#devtools #python #productivity

2 min read

soy

Mar 8

Reducing Token Consumption in Claude Code — FTS5 Knowledge DB + Tiered Index Design

#devtools #python #productivity

2 min read

soy

Mar 8

Implementing Stripe Checkout Billing in PatentLLM

#webdev #devops #infrastructure

2 min read

soy

Mar 8

Building a 5-in-1 App with Local LLM and Flutter

#ai #machinelearning #llm

2 min read

soy

Mar 8

Leveraging Claude Code's MCP Server

#devtools #python #productivity

2 min read

soy

Mar 8

LoRA and FT Are Unnecessary: How to Approach Distilled Models

#ai #machinelearning #llm

2 min read

soy

Mar 8

Lineage of OSS Supporting the AI Development Stack: Its Origins and Creators

#devtools #python #productivity

6 min read

soy

Mar 8

Running NVIDIA Nemotron-Nano-9B-v2-Japanese Locally: Mamba SSM + Thinking Mode Support

#ai #machinelearning #llm

2 min read

soy

Mar 8

Strategic Data Organization Techniques Using SQLite, JSONL, XML, and TSV: Lessons

#devtools #python #productivity

3 min read

soy

Mar 8

Shogi AI with RTX 5090 — Record of TensorRT FP8 Quantization and Floodgate Practical Games

#ai #gpu #performance

2 min read

soy

Mar 8

Practical Guide to Running Nemotron-Nano-9B-v2-Japanese with vLLM and Integrating it into Your Custom Application via an Open...

#ai #gpu #performance

6 min read

soy

Mar 8

Python Environment Management with uv: Introduction and Practical Use of a High-Speed Package Manager Replacing pip/venv

#devtools #python #productivity

3 min read

soy

Mar 8

Automatically Prevent Port Conflicts and Dangerous Commands Proactively with Claude Code's Hooks Feature

#devtools #python #productivity

2 min read

soy

Mar 8

Giving a 'Brain' to Minecraft NPCs with a Local LLM — Nemotron + Mineflayer Implementation Notes

#ai #machinelearning #llm

3 min read

soy

Mar 8

Exposing Multiple Web Applications from a Home Server with Cloudflare Tunnel + Caddy

#webdev #devops #infrastructure

2 min read

soy

Mar 8

Personal AI Development Environment Built with RTX 5090 + WSL2 — A Practical Setup Fully Utilizing 32GB GPU

#ai #gpu #performance

2 min read

soy

Mar 8

Individual Developer's Portfolio Strategy: Running 13 Projects on a Single RTX 5090

#ai #gpu #performance

2 min read

soy

Mar 8

Using Local LLMs as a "Batch Processing Engine" — A Design for Automatically Generating Artifacts from Your Own Data

#ai #machinelearning #llm

10 min read

soy

Mar 8

Fast Searching 4 Million Patent Records with FTS5

#ai #machinelearning #llm

2 min read

DEV Community

soy

Badges

Writing Debut

I Built a Free Patent Search Engine with 3.5M US Patents — No Login, Powered by SQLite FTS5

Want to connect with soy?

Operational Techniques for Automatically Starting vLLM, Flask, and cron with systemd Services in WSL2

Achieving Bidirectional Integration of Streamlit Backend Flutter Frontend in a WSL2 Environment

A Regulatory Analysis Dashboard for Fast Searching NITE CHRIP Data using FTS5

Searching Case Law PDFs with RAG — A Legal AI Search System using Gemini + SQLite FTS5

google-generativeai google-genai Migration Guide

Gemini 2.5 Flash x Nemotron 9B — Optimal Division of Roles for Cloud LLM and Local LLM

Reduce API Costs for Large-Scale Document Analysis with Gemini Context Caching

Skit: The Man Obsessed with Claude Code

Building a Free Research Agent with DuckDuckGo Search + Local LLM

A Daily Report System to Automatically Aggregate Claude Code + Gemini CLI Usage History Every Morning with Cron

Reducing Token Consumption in Claude Code — FTS5 Knowledge DB + Tiered Index Design

Implementing Stripe Checkout Billing in PatentLLM

Building a 5-in-1 App with Local LLM and Flutter

Leveraging Claude Code's MCP Server

LoRA and FT Are Unnecessary: How to Approach Distilled Models

Lineage of OSS Supporting the AI Development Stack: Its Origins and Creators

Running NVIDIA Nemotron-Nano-9B-v2-Japanese Locally: Mamba SSM + Thinking Mode Support

Strategic Data Organization Techniques Using SQLite, JSONL, XML, and TSV: Lessons

Shogi AI with RTX 5090 — Record of TensorRT FP8 Quantization and Floodgate Practical Games

Practical Guide to Running Nemotron-Nano-9B-v2-Japanese with vLLM and Integrating it into Your Custom Application via an Open...

Python Environment Management with uv: Introduction and Practical Use of a High-Speed Package Manager Replacing pip/venv

Automatically Prevent Port Conflicts and Dangerous Commands Proactively with Claude Code's Hooks Feature

Giving a 'Brain' to Minecraft NPCs with a Local LLM — Nemotron + Mineflayer Implementation Notes

Exposing Multiple Web Applications from a Home Server with Cloudflare Tunnel + Caddy

Personal AI Development Environment Built with RTX 5090 + WSL2 — A Practical Setup Fully Utilizing 32GB GPU

Individual Developer's Portfolio Strategy: Running 13 Projects on a Single RTX 5090

Using Local LLMs as a "Batch Processing Engine" — A Design for Automatically Generating Artifacts from Your Own Data

Fast Searching 4 Million Patent Records with FTS5