---
title: "Internals Decoded (@internalsdecoded) | Hashnode"
description: "AI Internals Explained Conversationally. Read the latest articles by Internals Decoded on Hashnode."
url: "https://hashnode.com/@internalsdecoded"
---

@internalsdecoded

# Internals Decoded

@internalsdecoded · Toronto, Canada · Joined August 2026

AI Internals Explained Conversationally

### [Entity Resolution: One Thing, Many Names](https://internals-decoded.hashnode.dev/entity-resolution?utm_source=hashnode&utm_medium=feed)

Sam K., Samuel Kim, sam@acme.com. Deciding what's the same thing is harder and more important than it looks.

14 min read · [internals-decoded.hashnode.dev](https://internals-decoded.hashnode.dev/entity-resolution?utm_source=hashnode&utm_medium=feed)

[2d ago](https://hashnode.com/posts/entity-resolution/6af3236ad0c1e8db1237a14d)

### [Debugging Agent Systems: Traces, Replays, Root Causes](https://internals-decoded.hashnode.dev/debugging-agent-systems?utm_source=hashnode&utm_medium=feed)

Agent failures raise no errors. The logging, tracing, and replay toolkit that makes them debuggable.

14 min read · [internals-decoded.hashnode.dev](https://internals-decoded.hashnode.dev/debugging-agent-systems?utm_source=hashnode&utm_medium=feed)

[5d ago](https://hashnode.com/posts/debugging-agent-systems/6aef2eed6f2a381cf4647352)

### [Evaluating Agents: Golden Sets, Judges, and Shadow Mode](https://internals-decoded.hashnode.dev/evaluating-ai-agents?utm_source=hashnode&utm_medium=feed)

Same input, different outputs. How to measure whether a nondeterministic system actually works, step by step as well as answer by answer.

17 min read · [internals-decoded.hashnode.dev](https://internals-decoded.hashnode.dev/evaluating-ai-agents?utm_source=hashnode&utm_medium=feed)

[Nov 4](https://hashnode.com/posts/evaluating-ai-agents/6aeb3a6dd76af20830d5f6e5)

### [Guardrails: Input, Process, Output](https://internals-decoded.hashnode.dev/agent-guardrails?utm_source=hashnode&utm_medium=feed)

Prompt injection, tool authorization, leak prevention: the three-layer safety system every agent needs, and the design that holds when a filter misses.

15 min read · [internals-decoded.hashnode.dev](https://internals-decoded.hashnode.dev/agent-guardrails?utm_source=hashnode&utm_medium=feed)

[Nov 1](https://hashnode.com/posts/agent-guardrails/6ae745ebfda6dd0f9e41c1f8)

### [Self-Improving Loops: Agents That Learn From Outcomes](https://internals-decoded.hashnode.dev/self-improving-loops?utm_source=hashnode&utm_medium=feed)

Track outcomes, mine the wins, update the prompts, guard every update: the shape of reinforcement learning without the gradients.

15 min read · [internals-decoded.hashnode.dev](https://internals-decoded.hashnode.dev/self-improving-loops?utm_source=hashnode&utm_medium=feed)

[Oct 29](https://hashnode.com/posts/self-improving-loops/6ae3435b3332c4401aa8309a)