---
title: "Karthik Karunanithi (@karthik006) | Hashnode"
description: "Solution Architect at IBM | Production Agentic AI, MCP, Enterprise Architecture. Read the latest articles by Karthik Karunanithi on Hashnode."
url: "https://hashnode.com/@karthik006"
---

@karthik006

# Karthik Karunanithi

@karthik006 · California · Joined April 2026

Solution Architect at IBM | Production Agentic AI, MCP, Enterprise Architecture

### [When an agent compacts its context, where does the evidence go?](https://karthikk.hashnode.dev/when-an-agent-compacts-its-context-where-does-the-evidence-go?utm_source=hashnode&utm_medium=feed)

Consider an agent running continuously for two hours. During this period, it handles files, utilizes tools, and steadily progresses. However, over time, it may forget previous actions, such as re-read

7 min read · [karthikk.hashnode.dev](https://karthikk.hashnode.dev/when-an-agent-compacts-its-context-where-does-the-evidence-go?utm_source=hashnode&utm_medium=feed)

[Sep 2](https://hashnode.com/posts/when-an-agent-compacts-its-context-where-does-the-evidence-go/6a9857d29be1c3bced297271)

### [Loop Engineering: Your Agent Doesn't Know When It's Done](https://karthikk.hashnode.dev/loop-engineering-your-agent-doesn-t-know-when-it-s-done?utm_source=hashnode&utm_medium=feed)

Everything held up in testing. That's how these stories always start. Then, in production, a downstream API timed out. Not at the beginning of the run, that would have been easy. It died in the middle

11 min read · [karthikk.hashnode.dev](https://karthikk.hashnode.dev/loop-engineering-your-agent-doesn-t-know-when-it-s-done?utm_source=hashnode&utm_medium=feed)

[Jul 26](https://hashnode.com/posts/loop-engineering-your-agent-doesn-t-know-when-it-s-done/6a6641e9c7d91d2e61a59188)

### [Harness Engineering: The Part of Agentic AI Nobody Writes About](https://karthikk.hashnode.dev/harness-engineering-the-part-of-agentic-ai-nobody-writes-about?utm_source=hashnode&utm_medium=feed)

Most production AI problems aren't the model or the prompt. They're the harness - feed context, manage tools, and handle failure. Here's what breaks.

13 min read · [karthikk.hashnode.dev](https://karthikk.hashnode.dev/harness-engineering-the-part-of-agentic-ai-nobody-writes-about?utm_source=hashnode&utm_medium=feed)

[2](https://hashnode.com/posts/harness-engineering-the-part-of-agentic-ai-nobody-writes-about/6a34dbbedfbba3127b16707e) [Jun 19](https://hashnode.com/posts/harness-engineering-the-part-of-agentic-ai-nobody-writes-about/6a34dbbedfbba3127b16707e)

### [Foundry vs Semantic Kernel vs AutoGen : What I Actually Use in Production](https://karthikk.hashnode.dev/foundry-vs-semantic-kernel-vs-autogen-what-i-actually-use-in-production?utm_source=hashnode&utm_medium=feed)

An opinionated comparison of Azure AI Foundry, Semantic Kernel, AutoGen based on what actually works in production enterprise agentic AI deployments.

7 min read · [karthikk.hashnode.dev](https://karthikk.hashnode.dev/foundry-vs-semantic-kernel-vs-autogen-what-i-actually-use-in-production?utm_source=hashnode&utm_medium=feed)

[May 20](https://hashnode.com/posts/foundry-vs-semantic-kernel-vs-autogen-what-i-actually-use-in-production/6a0d1bc88837277411c56d91)

### [95% of AI Pilots Fail. Here's What I See From the Inside.](https://karthikk.hashnode.dev/95-of-ai-pilots-fail-here-s-what-i-see-from-the-inside?utm_source=hashnode&utm_medium=feed)

Why most enterprise AI pilots never reach production, and the 5 patterns the teams that actually ship are getting right

6 min read · [karthikk.hashnode.dev](https://karthikk.hashnode.dev/95-of-ai-pilots-fail-here-s-what-i-see-from-the-inside?utm_source=hashnode&utm_medium=feed)

[May 5](https://hashnode.com/posts/95-of-ai-pilots-fail-here-s-what-i-see-from-the-inside/69f962d446610fd606582f09)