🌟 Editor's Note
Welcome to another exciting week of keeping it together with AI, so it doesn’t get the better of us.

Do you relate to the following? 👇

I used to spend 3+ hours every morning jumping between feeds, newsletters, and Slack threads trying to keep up with everything happening in AI, cloud, and security.

Useful? Sometimes. Overwhelming? Always. 🫣

I was there and frustrated with the 100s of tabs.

So I built Charlie, my AI assistant that reads everything I care about and sends me one clean Gmail email called Morning Brief every day at 6 AM.

Note - You don't need to know Coding, just a general understanding of the English Language and how to copy/paste information to follow this guide

TL;DR for Busy Readers

Want to try it yourself?

I’ve shared the workflow instructions, screenshot and starter kit -> 🔗 Read the Step-by-Step version (with Screenshots) + get the n8n workflow

🤖 Meet Charlie - My 6 AM AI Assistant

Charlie isn’t another chatbot. He’s a background process that is a part researcher, part editor, that curates what matters to me from the web, summarizes it in context, and drops it neatly into my inbox.

Each morning, Charlie:

  • Pulls the latest stories from my chosen sources (AI, Tech, Cloud, CyberSecurity, Men’s Fashion)

  • Uses Anthropic’s Claude to summarize and categorize them

  • Formats the summaries neatly inside Gmail as a DRAFT

  • Auto-Sends it to my Gmail 🙂

After a few days of fine-tuning, Charlie learned exactly how I like my brief: concise, relevant, and human.

🧱 Why I Built Charlie [Step-by-Step Guide]

I was tired of fragmented information. Every newsletter or feed seemed useful on its own, but together they created noise.

Charlie became a filter, a small system that condensed my digital environment into one calm signal. It’s the simplest automation I’ve built with the biggest mental ROI.

AI Workflow

⚙️ How to Build Your Own (in 5 Steps)

You’ll need:

  • n8n account (cloud or self-hosted)

  • Anthropic Claude (or OpenAI/Perplexity) API key

  • Tavily API (for web search)

  • Gmail account (for draft creation)

Then follow these 5 steps 👇

1️⃣ Spin up n8n and import the starter workflow (.JSON file in the AI Starter Kit below).

2️⃣ Add your credentials — API keys for Anthropic, Tavily, and Gmail.

3️⃣ Choose your sources — AI, cloud, tech news, etc.

4️⃣ Create the Gmail draft — Charlie formats it automatically.

5️⃣ Automate the run — schedule it daily at 6AM or weekly to your preference.

💡 Customise & Improve

Start small: 3 sources, 3 summaries. As you tune your brief, add more categories like “AI News,” “Tech,” or “Leadership.”

Over time, you can extend Charlie to:

  • Detecting duplicate stories via vector similarity

  • Personalizing summaries by topic depth or weekday

  • Send summaries to Slack or Teams instead of email

📦 Your AI Starter Kit for Charlie (Download ⬇️)

Want to try it?

I’ve shared the workflow instructions, screenshot and starter kit-> 🔗 Read the Step-by-Step version (with Screenshots) + get the n8n workflow

💬 Reflections

Charlie started as a tiny experiment and became a daily ritual that frees three hours of mental space each day.

This is just the beginning, i have already been working on friends of Charlie. 🙂.

Charlie has grown → My Entire AI Automation Suite (Friends of Charlie)

Today, Charlie is a lot more than just 1 Agent. More AI Agents Friends of Charlie will be released in future issues, subscribe if you would like to notified for the future AI automations.

Charlie & It's Friends

🗓️ Upcoming Deadlines

AI Automation Workshop (FREE)

Calling all early-stage startups: Submit your pitch for a chance to win $100,000 and investor spotlight.

  • Workshop Date: Dec 10, 2025

  • Location: Virtual

Building AI for Enterprise (Webinar)

A workshop on .

💡 Build Capability, Not Dependency

I help engineering and security leaders embed AI into their teams — turning existing talent into confident AI engineers.

Practical frameworks. Real workflows. No hype.

Did You Know? The first computer bug was literally a bug—in 1947, Grace Hopper found a moth trapped in a Harvard Mark II computer, coining the term "debugging" in the process.

Till next time,

Ashish Rajan

🧭 PS: If you enjoyed this post, consider subscribing to The Inference Stack, my newsletter where I share real AI workflows, frameworks, and experiments that help tech leaders stay ahead of the AI curve in their companies.

Keep reading

No posts found