John Wade

I'm building a knowledge management system across multiple LLM environments and documenting the failure modes that show up in sustained, multi-session work.

Joined on Feb 12, 2026

John Wade

Mar 8

When the Data Isn't Gone — It's Buried

#ai #architecture #knowledgemanagement #toolsforthought

8 min read

Want to connect with John Wade?

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

Create Account

Already have an account? Sign in

John Wade

Feb 26

When the Reasoning Doesn't Survive

#ai #architecture #knowledgemanagement #toolsforthought

4 min read

John Wade

Feb 22

When the Audit Contradicts Itself: Running a Knowledge Extraction Pipeline Under Load

#ai #architecture #knowledgemanagement #toolsforthought

7 min read

John Wade

Feb 21

When One Model Reviews Its Own Work: The Case for Adversarial Cross-Model Review

#ai #llm #testing #architecture

6 min read

John Wade

Feb 20

When a Session Forgets What It Knew: Why LLM State Management Breaks Under Load

#ai #llm #architecture #productivity

6 min read

John Wade

Feb 19

When a Model Goes Wide Instead of Deep: Installing Quality Gates That Actually Hold

#ai #llm #codequality #architecture

5 min read

John Wade

Feb 18

When the Sandbox Leaks: Context Contamination Across LLM Workspaces

#ai #llm #devops #architecture

5 min read

John Wade

Feb 16

When an AI Keeps Forgetting: Why LLM Workflows Collapse and What to Build Instead

#ai #llm #architecture #productivity

6 min read

John Wade

Feb 16

What Happens When You Try to Build Something Real with LLMs

#ai #llm #programming #toolsforthought

4 min read

DEV Community

John Wade

Badges

Writing Debut

When the Data Isn't Gone — It's Buried

Want to connect with John Wade?

When the Reasoning Doesn't Survive

When the Audit Contradicts Itself: Running a Knowledge Extraction Pipeline Under Load

When One Model Reviews Its Own Work: The Case for Adversarial Cross-Model Review

When a Session Forgets What It Knew: Why LLM State Management Breaks Under Load

When a Model Goes Wide Instead of Deep: Installing Quality Gates That Actually Hold

When the Sandbox Leaks: Context Contamination Across LLM Workspaces

When an AI Keeps Forgetting: Why LLM Workflows Collapse and What to Build Instead

What Happens When You Try to Build Something Real with LLMs