5 AI Automations That Run My Life in the Background

Photo by Alber on Unsplash

Years of automating repetitive tasks taught me one thing: “If your system still needs you, it’s not automated enough.” Let me explain.

A year ago, I was drowning in notifications, notes, and daily scripts. I had Python scripts for everything cleaning files, parsing emails, scraping data but they all needed me to run them. Manually.

That’s not automation. That’s procrastination disguised as productivity. Today, my setup runs 24/7 scheduling, summarizing, reminding, backing up, even learning. It’s like having an invisible assistant who doesn’t sleep, doesn’t complain, and doesn’t ask for coffee.

Let me show you five AI automations that quietly run my life in the background.

1. The Email Whisperer

Every morning, Gmail throws chaos at me receipts, newsletters, client updates, random “hey can you check this” messages. I built a small automation that uses OpenAI embeddings + Python to sort emails by intent.

Here’s the trick: instead of keyword filters, I classify each message semantically.

from openai import OpenAI
import imaplib, email
client = OpenAI()
mail = imaplib.IMAP4_SSL("imap.gmail.com")
mail.login("[email protected]", "password")
mail.select("inbox")
_, data = mail.search(None, "ALL")
for num in data[0].split():
    _, msg_data = mail.fetch(num, "(RFC822)")
    msg = email.message_from_bytes(msg_data[0][1])
    content = msg.get_payload(decode=True).decode("utf-8")
    
    category = client.embeddings.create(input=content)
    # Compare embedding with predefined vectors: 'work', 'personal', 'spam', etc.This script feeds the text into embeddings and matches them against a few predefined categories. The result?
 My emails land neatly into “Work,” “Personal,” “Action Needed,” and “Spam-but-not-quite.”

Pro Tip: 
Keywords are static. Embeddings understand context. That’s a huge difference.

2. My Daily Summary Bot

Every night at 11:00 PM, I get a Slack message: “Zaki, here’s what happened today.” It summarizes everything from emails and tasks to Git commits. It’s basically my day condensed into 6 bullet points.

Here’s how it works:

A Python cron job pulls data from Notion, Gmail, and GitHub APIs. It concatenates summaries into a single text blob. GPT condenses it using this simple prompt:

prompt = f"""
Summarize this day for me in 6 bullet points.
Highlight key tasks and outcomes.Data:
{combined_data}
"""

The best part? I don’t even open Slack. My phone buzzes, and I know what I achieved. Quote I love: “You can’t improve what you don’t measure.” Peter Drucker Automation isn’t about laziness it’s about awareness without effort.

3. My Content Pipeline (Zero Clicks Needed)

As a writer, I get random bursts of ideas usually when I’m not near my laptop. So I automated it.

Whenever I send a WhatsApp message to myself starting with /idea, it triggers a webhook that stores it in Notion, runs it through GPT to generate a potential outline, and tags it with a topic category.

if message.startswith("/idea"):
    idea = message.replace("/idea", "").strip()
    outline = client.responses.create(
        model="gpt-4.1",
        input=f"Generate a Medium article outline for: {idea}"
    )
    save_to_notion(idea, outline.output_text)

I wake up to a list of structured article ideas instead of messy notes. It’s like turning my stream of consciousness into publish-ready material. Pro Tip:
Don’t rely on motivation. Rely on systems.

4. My System Cleaner (The Silent Janitor)

Here’s a confession: I used to hoard files. Logs, datasets, screenshots, backups you name it. My system looked like a digital landfill. So I wrote a daemon that auto-cleans unused files and large folders weekly. It even archives old projects into cloud storage using simple rules:

import os, shutil, datetime
def clean_old_files(path, days=30):
    now = datetime.datetime.now()
    for file in os.listdir(path):
        full_path = os.path.join(path, file)
        if os.path.isfile(full_path):
            last_access = datetime.datetime.fromtimestamp(os.path.getatime(full_path))
            if (now - last_access).days > days:
                shutil.move(full_path, "/archive")

Result: Disk space recovered: 128GB Random “where did that file go” moments: 0 Peace of mind: Infinite

5. The Brain That Learns While I Sleep

This one’s my favorite. I have a script that monitors the topics I consume videos watched, articles read, and code searched. It feeds that data into GPT weekly to recommend 3 new learning topics.

Example:
If I’ve been reading about “LangChain” and “vector databases,” it might suggest “embedding compression techniques.”

That’s personalized learning on autopilot. It’s scary how well it works. Sometimes it feels like my AI knows what I’ll study before I do.

Automation tip: 
Combine AI models with your own data history that’s where the real power is.

The Bigger Picture

These automations don’t make me a robot.
They make me human again free from the endless clicking, filtering, and remembering.

Because here’s the truth:

“Automation doesn’t replace effort. It reallocates it to where it actually matters.” Before I automated, I was busy.
After automation, I’m effective.

If you take one thing from this article, let it be this: 
Don’t automate to save time. Automate to save attention. The moment your system runs without you, you’re no longer working for your tools they’re working for you.

Written by Zaki Ali (Age 18) on a system that auto-backed up, formatted, and posted this draft while he was making coffee.

read the full article here: https://ai.plainenglish.io/5-ai-automations-that-run-my-life-in-the-background-8176df06aadb