IC4 Weekly #01 — Signals from the Week

Intro

Hello Everyone,

I’m Madhusudhan Reddy, K — a Staff Software Engineer at ServiceNow with almost 11 years of building things end-to-end—from frontends and backends to SDKs and everything in between.

Part of my job as a Staff Engineer is staying curious—reading, exploring, watching podcasts, talking to people etc. But most of what I consume fades within days. Not because it wasn’t valuable, but because I never paused to capture or share it.

This newsletter is me fixing that. Every week, I write down the few things that actually stayed with me from that week—an article that shifted my thinking, a tool I wish I’d found earlier, an idea I keep coming back to. I’ll share what it is, why it stuck, and where to find it. Nothing more.

Tools & Tech I Looked At

1. SimAi

Sim.ai — An open-source, visual platform for building AI agent workflows. Think drag-and-drop automation, but designed with agents in mind. It sits somewhere between n8n/Zapier and fully custom agent code, with support for local models via Ollama.

Feels useful for quick experiments and internal tools.

2. MDX

MDX — Markdown with JSX embedded. The main win here is reducing friction between content and code—especially for docs, blogs, and dev-facing content. You write mostly Markdown, drop in components where needed, and keep everything readable.

Not new, but still underrated if you’re building documentation-heavy products.

Playground

3. HTMX

A small library (~14KB) that lets you build interactive apps using server-rendered HTML and simple attributes—no heavy frontend framework required.

Motivation behind this lib:

Why should only <a> & <form> be able to make HTTP requests?
Why should only click & submit events trigger them?
Why should only GET & POST methods be available?
Why should you only be able to replace the entire screen?

By removing these constraints, htmx completes HTML as a hypertext.

What’s interesting isn’t performance alone, but the shift in mindset: push complexity back to the server, keep the client dumb. For CRUD apps, dashboards, and internal tools, this approach often makes more sense than reaching for React by default.

Useful Public Repositories

AGENTS.md — A simple, open format for guiding coding agents

Think of AGENTS.md as a README for agents: a dedicated, predictable place to provide context and instructions to help AI coding agents work on your project.

Productivity Concepts

While reading through a book I came across couple of productivity strategies.

1. SNR - Signal to Noise Ratio (80/20)

Signal-to-noise ratio (SNR or S/N) is a measure used in science and engineering that compares the level of a desired signal to the level of background noise.

But SNR is used to measure productivity as well. It is basically the ratio between the signal (things to be done immediately which are crucial, not good to have but crucial) and noise (everything else). The higher SNR is, the more productive a person is.

2. Pareto Principle (20/80)

This is similar to SNR. It measures how a small set of actions usually creates most of the outcome. This applies to learning, coding, and management work. It’s not a law—but it’s a good starting point when you’re overwhelmed and need to prioritize fast.

Examples:

20% of your team will do 80% of work
20% of your stories deliver 80% of product
20% of code is responsible for 80% of defects :)
20% of conversations builds 80% of trust in a relationship

We can apply this for almost everything in life, and this does not solve the problem by itself but it gives us a hint on which 20% / 80% we need to look at.

3. Strategic Ignorance

Practice strategic ignorance. Not everything deserves your attention, even if it seems useful. The more you try to stay on top of, the harder it becomes to focus on what actually matters. Protect your signal-to-noise ratio.

You don’t need to follow every tech drama on X, read every trending article, or have an opinion on every new framework. More input doesn’t mean more clarity. Often, it just drowns the signal in noise and fragments your focus.

Examples:

Ignoring Slack/Teams channels that are interesting but irrelevant to your work
Creating automation rules to move noisy mails
Unsubscribing from newsletters you save but never read
Choosing not to learn a tool until you actually need it

People, Concepts & AI

Geoffrey Hinton

I recently picked up Genius Makers: The Mavericks Who Brought AI to Google, Facebook and the World by Cade Metz. And learned about his crucial role in bringing current genAI into the world. One of the core figures behind modern deep learning—and now one of the loudest voices warning about AI risks.

Links Worth Bookmarking

Sim.ai — Open-source AI workflow builder
MDX — Markdown + JSX
HTMX — Server-driven interactivity
AGENTS.md — Instructions format for coding agents

(Links included so you can dig deeper if something clicks.)