menu

dechive

@Dechive

Dechive

Practical notes on AI verification, prompting, structured output, workflows, and...
Korea dechive.dev Joined May 2026
1.7k Points • 31 Badges • 7 Connections • 7 Followers • 9 Following
Profile Stats Wall Posts Badges

About

Dechive is an AI verification archive for developers, builders, and solo founders.

I write practical notes about prompt engineering, RAG, structured output, hallucination, AI work... Show more

Language & Tools

JavaScript, React, Next.js, Supabase, SQL, Markdown, GitHub Actions, RAG, embeddings, structured output, AI-assisted development, Claude Code, Codex, ChatGPT, Gemini.

Currently Exploring

AI workflows, prompt engineering, RAG, structured output, MCP, n8n, solo SaaS, knowledge systems, and tools that help people think, verify, and build with better direction.

Achievements

Built and maintain Dechive, an AI verification archive for understanding, testing, and trusting AI-generated answers more carefully.

It includes archive pages, deep-dive essays, SEO metadata, translation workflows, and an embedding-based knowledge system.

Fun Fact

I think of Dechive less as a blog and more as a verification archive.

Each post starts from a single question. Some become short records; others become deep dives that organize a concept carefully before we trust it.

Random Dev Quote

Before asking AI for an answer, hand over your thinking first.

Top Tags

User Activities

JulAugSepOctNovDecJanFebMarAprMayJun
Mon
Tue
Wed
Thu
Fri
Sat
Sun
Less More

Contributions
Joined: 2 months (since May 11)
Extra privileges: Editing any comment
Full Name: Dechive
Headline: Practical notes on AI verification, prompting, structured output, workflows, and developer judgment.
About: Dechive is an AI verification archive for developers, builders, and solo founders.

I write practical notes about prompt engineering, RAG, structured output, hallucination, AI workflows, and the human judgment needed before we trust generated answers.

Core idea:
AI creates answers. Dechive verifies them.

Main archive: https://dechive.dev
Location: Korea
Website: https://dechive.dev
Languges & Tools: JavaScript, React, Next.js, Supabase, SQL, Markdown, GitHub Actions, RAG, embeddings, structured output, AI-assisted development, Claude Code, Codex, ChatGPT, Gemini.
Currently Exploring: AI workflows, prompt engineering, RAG, structured output, MCP, n8n, solo SaaS, knowledge systems, and tools that help people think, verify, and build with better direction.
Achievements: Built and maintain Dechive, an AI verification archive for understanding, testing, and trusting AI-generated answers more carefully.

It includes archive pages, deep-dive essays, SEO metadata, translation workflows, and an embedding-based knowledge system.
Fun Fact: I think of Dechive less as a blog and more as a verification archive.

Each post starts from a single question. Some become short records; others become deep dives that organize a concept carefully before we trust it.
Random Dev Quote: Before asking AI for an answer, hand over your thinking first.
Latest Video:
chevron_left

Active in these Groups:

AI
1199 members
Open Source
1133 members
WebDev
950 members

Latest Jobs

View all jobs →