DEV Community

juan gonzalez profile picture

juan gonzalez

I build verification systems for AI agents and automation: claims, evidence, verdicts, and fail-closed execution when it matters. Founder of TaiwildLab.

Location Alicante, Spain Joined Joined on  Personal website https://taiwildlab.com github website twitter website

Education

Self-taught developer

Work

Founder @ TaiwildLab

The Cost of Trusting the Wrong Claim in AI Systems

The Cost of Trusting the Wrong Claim in AI Systems

Comments
1 min read

Want to connect with juan gonzalez?

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

Already have an account? Sign in
# Don’t Trust the Agent’s “Done”: Verify the System State

# Don’t Trust the Agent’s “Done”: Verify the System State

Comments
4 min read
Why an Evolutionary System Needs Memory

Why an Evolutionary System Needs Memory

Comments
2 min read
I shipped a fix. The system recovered. The fix wasn't why.

I shipped a fix. The system recovered. The fix wasn't why.

Comments
3 min read
A Shadow Decision System Found a Semantic Boundary Bug in 14 Minutes

A Shadow Decision System Found a Semantic Boundary Bug in 14 Minutes

Comments
2 min read
La confianza no es una propiedad del agente. Es una propiedad de la arquitectura.

La confianza no es una propiedad del agente. Es una propiedad de la arquitectura.

Comments
1 min read
I Run 20+ Autonomous Trading Agents. Here Are the 5 Memory Failures That Almost Killed the System.

I Run 20+ Autonomous Trading Agents. Here Are the 5 Memory Failures That Almost Killed the System.

Comments
1 min read
loading...