Forem

Kwansub Yun profile picture

Kwansub Yun

Scientific AI Systems Architect | BioAI & Physics-ML | Deploying Governance-Ready AI Systems to Production

Location Yala, Thailand Joined Joined onย  Personal website https://flamehaven.space github website twitter website

Work

Founder & Scientific AI Systems Architect โ€” Flamehaven

How do you know when your entire AI pipeline is wrong โ€” not just one model? (EXP-033)
Cover image for How do you know when your entire AI pipeline is wrong โ€” not just one model? (EXP-033)

How do you know when your entire AI pipeline is wrong โ€” not just one model? (EXP-033)

Comments
6 min read

Want to connect with Kwansub Yun?

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

Already have an account? Sign in
I Built an Ecosystem of 46 AI-Assisted Repos. Then I Realized It Might Be Eating Itself.
Cover image for I Built an Ecosystem of 46 AI-Assisted Repos. Then I Realized It Might Be Eating Itself.

I Built an Ecosystem of 46 AI-Assisted Repos. Then I Realized It Might Be Eating Itself.

Comments
10 min read
The Tool That Turned on Itself: AI-Slop-Detector v2.9.0 v2.9.1
Cover image for The Tool That Turned on Itself: AI-Slop-Detector v2.9.0 v2.9.1

The Tool That Turned on Itself: AI-Slop-Detector v2.9.0 v2.9.1

Comments
12 min read
We Accidentally Rewarded AI Spaghetti Code. Here is the Math We Used to Fix It. : AI-Slop Detector v2.8.0
Cover image for We Accidentally Rewarded AI Spaghetti Code. Here is the Math We Used to Fix It. : AI-Slop Detector v2.8.0

We Accidentally Rewarded AI Spaghetti Code. Here is the Math We Used to Fix It. : AI-Slop Detector v2.8.0

1
Comments
5 min read
I Built a Physics Verification Engine with Google Gemini. Here's What 700 Tests and 12 Months Taught Me.

Built with Google Gemini: Writing Challenge

I Built a Physics Verification Engine with Google Gemini. Here's What 700 Tests and 12 Months Taught Me.

2
Comments
16 min read
What an AI Reasoning Engine Built for Alzheimer's Metabolic Research: A Code Walkthrough
Cover image for What an AI Reasoning Engine Built for Alzheimer's Metabolic Research: A Code Walkthrough

What an AI Reasoning Engine Built for Alzheimer's Metabolic Research: A Code Walkthrough

Comments
6 min read
I Survived a 300-Year Flood. Then I Built Flood Ready: On-Device AI for When the Internet Fails
Cover image for I Survived a 300-Year Flood. Then I Built Flood Ready: On-Device AI for When the Internet Fails

DEV Weekend Challenge: Community

I Survived a 300-Year Flood. Then I Built Flood Ready: On-Device AI for When the Internet Fails

2
Comments 3
11 min read
Your Agentic Stack Has Two Layers. It Needs Three.
Cover image for Your Agentic Stack Has Two Layers. It Needs Three.

Your Agentic Stack Has Two Layers. It Needs Three.

1
Comments 2
15 min read
From Fail-Closed Blocking to Reproducible PASS/BLOCK Separation (EXP-032B)
Cover image for From Fail-Closed Blocking to Reproducible PASS/BLOCK Separation (EXP-032B)

From Fail-Closed Blocking to Reproducible PASS/BLOCK Separation (EXP-032B)

Comments
13 min read
Chaos Engineering for AI: Validating a Fail-Closed Pipeline with Fake Data and Math
Cover image for Chaos Engineering for AI: Validating a Fail-Closed Pipeline with Fake Data and Math

Chaos Engineering for AI: Validating a Fail-Closed Pipeline with Fake Data and Math

Comments
4 min read
From 97% Model Accuracy to 74% Clinical Reliability: Building RSN-NNSL-GATE-001
Cover image for From 97% Model Accuracy to 74% Clinical Reliability: Building RSN-NNSL-GATE-001

From 97% Model Accuracy to 74% Clinical Reliability: Building RSN-NNSL-GATE-001

Comments
7 min read
Copilot Guardian - Beyond the Red X: AI That Debugs With You, Not For You
Cover image for Copilot Guardian - Beyond the Red X: AI That Debugs With You, Not For You

GitHub Copilot CLI Challenge Submission

Copilot Guardian - Beyond the Red X: AI That Debugs With You, Not For You

7
Comments
6 min read
AI-SLOP Detector v2.7.0 โ€” Why We Built a Linter We Actually Use
Cover image for AI-SLOP Detector v2.7.0 โ€” Why We Built a Linter We Actually Use

AI-SLOP Detector v2.7.0 โ€” Why We Built a Linter We Actually Use

Comments
2 min read
Trinity Protocol Part 2: When Adding Chai-1 and Boltz-2 Exposed Hidden Model Disagreement
Cover image for Trinity Protocol Part 2: When Adding Chai-1 and Boltz-2 Exposed Hidden Model Disagreement

Trinity Protocol Part 2: When Adding Chai-1 and Boltz-2 Exposed Hidden Model Disagreement

Comments
10 min read
Orchestrating AlphaFold 3 & 2 with Python: Handling AI Hallucinations using Adapter Pattern
Cover image for Orchestrating AlphaFold 3 & 2 with Python: Handling AI Hallucinations using Adapter Pattern

Orchestrating AlphaFold 3 & 2 with Python: Handling AI Hallucinations using Adapter Pattern

Comments
6 min read
I Integrated AlphaFolder3 & AlphaGenome. It Looked Perfect. Then It Failed the "Honesty Test."
Cover image for I Integrated AlphaFolder3 & AlphaGenome. It Looked Perfect. Then It Failed the "Honesty Test."

I Integrated AlphaFolder3 & AlphaGenome. It Looked Perfect. Then It Failed the "Honesty Test."

Comments
6 min read
Why Reasoning Models Die in Production (and the Test Harness I Ship Now)

Why Reasoning Models Die in Production (and the Test Harness I Ship Now)

Comments
5 min read
How Failing in 2 Hours Saved 8 Months of Drug R&D: Engineering a "Truthful Null" with Upadacitinib
Cover image for How Failing in 2 Hours Saved 8 Months of Drug R&D: Engineering a "Truthful Null" with Upadacitinib

How Failing in 2 Hours Saved 8 Months of Drug R&D: Engineering a "Truthful Null" with Upadacitinib

Comments 1
6 min read
RExSyn Nexus 0.6.1 - Stop Hallucinating Proteins: How We Built a 7D Reasoning Engine with AlphaFold3
Cover image for RExSyn Nexus 0.6.1 - Stop Hallucinating Proteins: How We Built a 7D Reasoning Engine with AlphaFold3

RExSyn Nexus 0.6.1 - Stop Hallucinating Proteins: How We Built a 7D Reasoning Engine with AlphaFold3

Comments
7 min read
Implementing "Refusal-First" RAG: Why We Architected Our AI to Say 'I Don't Know'
Cover image for Implementing "Refusal-First" RAG: Why We Architected Our AI to Say 'I Don't Know'

Implementing "Refusal-First" RAG: Why We Architected Our AI to Say 'I Don't Know'

Comments
3 min read
LOGOS LawBinder: From Governed Reasoning to Audit-Grade Execution
Cover image for LOGOS LawBinder: From Governed Reasoning to Audit-Grade Execution

LOGOS LawBinder: From Governed Reasoning to Audit-Grade Execution

Comments
2 min read
LOGOS v1.4.1: Building Multi-Engine AI Reasoning You Can Actually Trust

LOGOS v1.4.1: Building Multi-Engine AI Reasoning You Can Actually Trust

Comments
2 min read
LawBinder v1.3.0: Governance as a Kernel (Not a Guardrail)

LawBinder v1.3.0: Governance as a Kernel (Not a Guardrail)

Comments
4 min read
๐Ÿงฉ AI Slop Detector v2.6.3 is live โ€” now on VS Code
Cover image for ๐Ÿงฉ AI Slop Detector v2.6.3 is live โ€” now on VS Code

๐Ÿงฉ AI Slop Detector v2.6.3 is live โ€” now on VS Code

Comments 1
3 min read
AI Slop Detector v2.6.2: Integration Test Evidence (because โ€œgreen CIโ€ can still be hollow)
Cover image for AI Slop Detector v2.6.2: Integration Test Evidence (because โ€œgreen CIโ€ can still be hollow)

AI Slop Detector v2.6.2: Integration Test Evidence (because โ€œgreen CIโ€ can still be hollow)

Comments
2 min read
Iโ€™m Not Building AI Demos. Iโ€™m Building AI Audits (ASDP + Slop Gates)
Cover image for Iโ€™m Not Building AI Demos. Iโ€™m Building AI Audits (ASDP + Slop Gates)

Iโ€™m Not Building AI Demos. Iโ€™m Building AI Audits (ASDP + Slop Gates)

Comments
3 min read
AI Slop Detector v2.6.1 Self-Audit: I planted 3 bad files โ€” did it catch them?
Cover image for AI Slop Detector v2.6.1 Self-Audit: I planted 3 bad files โ€” did it catch them?

AI Slop Detector v2.6.1 Self-Audit: I planted 3 bad files โ€” did it catch them?

Comments
3 min read
AI-SLOP Detector v2.6.1: It now audits itself (and thatโ€™s the point)
Cover image for AI-SLOP Detector v2.6.1: It now audits itself (and thatโ€™s the point)

AI-SLOP Detector v2.6.1: It now audits itself (and thatโ€™s the point)

Comments
3 min read
๐—ช๐—ต๐˜† ๐—”๐—œ-๐—š๐—ฒ๐—ป๐—ฒ๐—ฟ๐—ฎ๐˜๐—ฒ๐—ฑ ๐—–๐—ผ๐—ฑ๐—ฒ ๐—ข๐—ณ๐˜๐—ฒ๐—ป ๐—Ÿ๐—ผ๐—ผ๐—ธ๐˜€ โ€œ๐—–๐—ผ๐—บ๐—ฝ๐—น๐—ฒ๐˜๐—ฒโ€ โ€” ๐—ฏ๐˜‚๐˜ ๐—œ๐˜€๐—ปโ€™๐˜โ€”๐—ฎ๐—ป๐—ฑ ๐˜„๐—ต๐˜† ๐—œ ๐—ฏ๐˜‚๐—ถ๐—น๐˜ ๐—”๐—œ-๐—ฆ๐—Ÿ๐—ข๐—ฃ ๐——๐—ฒ๐˜๐—ฒ๐—ฐ๐˜๐—ผ๐—ฟ
Cover image for ๐—ช๐—ต๐˜† ๐—”๐—œ-๐—š๐—ฒ๐—ป๐—ฒ๐—ฟ๐—ฎ๐˜๐—ฒ๐—ฑ ๐—–๐—ผ๐—ฑ๐—ฒ ๐—ข๐—ณ๐˜๐—ฒ๐—ป ๐—Ÿ๐—ผ๐—ผ๐—ธ๐˜€ โ€œ๐—–๐—ผ๐—บ๐—ฝ๐—น๐—ฒ๐˜๐—ฒโ€ โ€” ๐—ฏ๐˜‚๐˜ ๐—œ๐˜€๐—ปโ€™๐˜โ€”๐—ฎ๐—ป๐—ฑ ๐˜„๐—ต๐˜† ๐—œ ๐—ฏ๐˜‚๐—ถ๐—น๐˜ ๐—”๐—œ-๐—ฆ๐—Ÿ๐—ข๐—ฃ ๐——๐—ฒ๐˜๐—ฒ๐—ฐ๐˜๐—ผ๐—ฟ

๐—ช๐—ต๐˜† ๐—”๐—œ-๐—š๐—ฒ๐—ป๐—ฒ๐—ฟ๐—ฎ๐˜๐—ฒ๐—ฑ ๐—–๐—ผ๐—ฑ๐—ฒ ๐—ข๐—ณ๐˜๐—ฒ๐—ป ๐—Ÿ๐—ผ๐—ผ๐—ธ๐˜€ โ€œ๐—–๐—ผ๐—บ๐—ฝ๐—น๐—ฒ๐˜๐—ฒโ€ โ€” ๐—ฏ๐˜‚๐˜ ๐—œ๐˜€๐—ปโ€™๐˜โ€”๐—ฎ๐—ป๐—ฑ ๐˜„๐—ต๐˜† ๐—œ ๐—ฏ๐˜‚๐—ถ๐—น๐˜ ๐—”๐—œ-๐—ฆ๐—Ÿ๐—ข๐—ฃ ๐——๐—ฒ๐˜๐—ฒ๐—ฐ๐˜๐—ผ๐—ฟ

Comments 3
2 min read
๐Ÿฆโ€๐Ÿ”ฅ Weekly Flamehaven Patch Report โ€” this week was a โ€œstack alignmentโ€ week.

๐Ÿฆโ€๐Ÿ”ฅ Weekly Flamehaven Patch Report โ€” this week was a โ€œstack alignmentโ€ week.

Comments
2 min read
HRPO-X v1.0.1: from HRPO paper production-hardened runnable code

HRPO-X v1.0.1: from HRPO paper production-hardened runnable code

Comments
2 min read
Romance Is Dead: Open Sourceโ€™s Critical Inflection Point and the 14,000,605-to-1 Survival Strategy

Romance Is Dead: Open Sourceโ€™s Critical Inflection Point and the 14,000,605-to-1 Survival Strategy

Comments
4 min read
Undo Beats IQ: Building Flamehaven as a Governed AI Runtime (Not a Prompt App)

Undo Beats IQ: Building Flamehaven as a Governed AI Runtime (Not a Prompt App)

Comments
2 min read
Turning a Research Paper into a Runnable System
Cover image for Turning a Research Paper into a Runnable System

Turning a Research Paper into a Runnable System

Comments
2 min read
๐—˜๐—ป๐˜๐—ฒ๐—ฟ๐—ฝ๐—ฟ๐—ถ๐˜€๐—ฒ ๐—Ÿ๐—Ÿ๐—  ๐˜€๐˜†๐˜€๐˜๐—ฒ๐—บ๐˜€ ๐—ต๐—ถ๐˜ ๐—ฎ ๐—ฝ๐—ฟ๐—ฎ๐—ฐ๐˜๐—ถ๐—ฐ๐—ฎ๐—น ๐˜„๐—ฎ๐—น๐—น: ๐˜๐—ต๐—ฒ ๐—บ๐—ผ๐—ฟ๐—ฒ ๐˜†๐—ผ๐˜‚ ๐˜ƒ๐—ฒ๐—ฟ๐—ถ๐—ณ๐˜†, ๐˜๐—ต๐—ฒ ๐˜€๐—น๐—ผ๐˜„๐—ฒ๐—ฟ ๐˜‚๐˜€๐—ฒ๐—ฟ๐˜€ ๐˜„๐—ฎ๐—ถ๐˜.

๐—˜๐—ป๐˜๐—ฒ๐—ฟ๐—ฝ๐—ฟ๐—ถ๐˜€๐—ฒ ๐—Ÿ๐—Ÿ๐—  ๐˜€๐˜†๐˜€๐˜๐—ฒ๐—บ๐˜€ ๐—ต๐—ถ๐˜ ๐—ฎ ๐—ฝ๐—ฟ๐—ฎ๐—ฐ๐˜๐—ถ๐—ฐ๐—ฎ๐—น ๐˜„๐—ฎ๐—น๐—น: ๐˜๐—ต๐—ฒ ๐—บ๐—ผ๐—ฟ๐—ฒ ๐˜†๐—ผ๐˜‚ ๐˜ƒ๐—ฒ๐—ฟ๐—ถ๐—ณ๐˜†, ๐˜๐—ต๐—ฒ ๐˜€๐—น๐—ผ๐˜„๐—ฒ๐—ฟ ๐˜‚๐˜€๐—ฒ๐—ฟ๐˜€ ๐˜„๐—ฎ๐—ถ๐˜.

Comments
3 min read
loading...