Autom8wmark

23/05/2026

Currently building a custom n8n automation project that fully streamlines user offboarding for an MSP. The goal? Eliminate human error, revoke access instantly, and save hours of manual work.

The screenshot below shows the initial workflows capturing all the critical data needed to kick off the process smoothly.

👇 Need to automate your own MSP workflows or internal operations?
Drop a comment below or send me a DM to see how we can optimize your business!

24/04/2026

🚀 Automation Spotlight: From Quote to Payment-Ready Instantly

Still manually setting up customers in your payment gateway every time you send a quote? That delay can cost you momentum… and sometimes the sale.

Here’s how we solved that 👇

We built an end-to-end automation that connects Quoter directly to Alternative Payments using n8n so the moment a quote is published:

✅ The recipient is automatically checked in your payment system.
✅ If they already exist → nothing breaks, nothing duplicates.
✅ If they don’t → they’re instantly created under the correct company.
✅ No manual approvals. No back-and-forth. No delays.

💡 What this means for your business:
• Faster customer onboarding
• Smoother payment experience
• Reduced manual workload for your team
• Higher chances of closing deals while interest is hot

Built with: Webhooks, OAuth 2.0, smart filtering, and conditional logic and all orchestrated seamlessly in n8n.

This isn’t just automation. It’s removing friction from your revenue flow.

If you’re still juggling quotes and payment setups manually, it might be time to rethink your workflow.

💬 Curious how this could work for your setup? Let’s talk.

29/03/2026

🚀 Automation Project Complete: Smarter Device Ownership in ConnectWise

After weeks of building, testing, and refining, I’ve just completed an automation workflow that solves a very common problem in MSP environments using ConnectWise and ImmyBot.

💡 The Problem
Device ownership in ConnectWise often becomes outdated over time.
Users change roles, devices get reassigned, and suddenly you’re not sure who actually owns what.

Manually fixing this across dozens of companies?
Time-consuming and error-prone.

⚙️ The Solution
Built an automated workflow (using n8n) that:

✅ Reviews devices across companies.
✅ Cross-checks actual device usage from ImmyBot
✅ Detects mismatches in assigned users
✅ Automatically updates ConnectWise with the correct contact
✅ Sends daily reports via Microsoft Teams

🧠 Smart Matching Logic
To ensure accuracy, the workflow uses:

Email matching
Name verification
Fallback name search

Handling real-world edge cases like mismatched usernames and emails.

🖥️ Device Handling

Windows workstations → assigned to real primary user
Macs → skipped (not managed in ImmyBot)
Servers / network devices → assigned to “Proactive Courant”

📊 Daily Visibility
Every run provides a summary:

Total updates made
Any failures
Full list of updated devices and users

🔥 Impact

Saves hours of manual work
Improves data accuracy
Runs automatically every day
Reduces confusion during support

✅ Status:
Fully built, tested across all companies, and ready for deployment.

This was a great example of how automation can clean up real operational problems in MSP environments. Looking forward to building more solutions like this.

27/02/2026

I've been learning deeper into AI/LLM Engineering and been working with Gradio to create a nice UI for user experience. In the screenshot, you can see a simple input from user and output which is a callback to an LLM. You can also select which model you want to use and it will display it's output in a nice stream and markdown.

From here, you can improve the UI and the functions to create a chatbot similar to ChatGPT. I think.

Let me know your thoughts or feedback in the comments!

18/02/2026

I been learning AI/LLM Engineering recently on training vs inference-time scaling, and it was surprisingly interesting.

I asked a very simple probability puzzle:

“You toss 2 coins. One of them is heads. What’s the probability the other is tails?”

Then I changed just one thing: the reasoning_effort parameter.

Same prompt.
Same model family.
Only difference was:

reasoning_effort="minimal"

reasoning_effort="low"

That’s it.

And the outputs changed.

This is inference-time scaling in action.

Instead of retraining a bigger model or adding more data, you’re simply adjusting how much “thinking effort” the model uses when answering.

Think of it like this:

Minimal reasoning_effort → quick, fast, low compute answer

Higher reasoning_effort → more deliberate reasoning before responding

It’s like giving the model either:

5 seconds to answer

or a bit more time and scratch paper

For years, AI progress was mostly about training-time scaling: bigger models, more GPUs, more data.

But this experiment shows something subtle and powerful:
You can change behavior and reliability just by changing how much compute the model uses at inference time.

That’s a big shift.

We’re moving from:
“How big is the model?”

To:
“How much effort does it use when thinking?”

And that might end up being just as important.

17/02/2026

📅 No More Double Bookings. No More Manual Scheduling.

Still replying to messages like:
“Is 9:00 AM available?”
“Sorry, that slot is already taken.”
“Can we move it to 9:30 instead?”

I just built an automated appointment booking system that:

✅ Lets clients choose their preferred date and time
✅ Instantly checks real-time calendar availability
✅ Prevents double booking
✅ Automatically confirms the appointment
✅ Sends email notifications to the client
✅ Updates the business calendar automatically

All in seconds. No back-and-forth. No manual checking.

Perfect for:
Clinics
Consultants
Coaches
Freelancers
Service-based businesses

Imagine your clients booking themselves while you focus on actual work.
If you're still manually managing appointments, you're wasting valuable time every week.
Automation isn’t about replacing people, It’s about removing repetitive tasks so you can grow.

If you want something like this for your business, send me a message.

01/02/2026

The difference in response with a single word change in system prompt. 🤣🤣🤣

30/01/2026

I’ve been helping businesses remove manual work using automation—and this is one of my favorite builds so far 👇

A client scheduling system that:
• Accepts bookings online
• Confirms appointments automatically
• Keeps everything synced and organized

Less admin work. Fewer mistakes. Happier clients.

If you want something similar for your business, feel free to message me. Happy to explain how it works (in simple terms 😊)

13/01/2026

Just wrapped up Phase 1 of an AI-powered system I’m building for a client on Fiverr and it’s designed to solve a problem a lot of businesses deal with every day:

Too much information, not enough time to make sense of it all.

What makes this especially powerful is that it can ingest an existing internal knowledge base. Things like emails, documents, and past conversations and turn all of that into a smart, searchable system that an AI assistant can understand.

In real life, that means:
✔ Less time digging through emails and files
✔ Faster decisions
✔ More organized teams
✔ Less stress
✔ Better workflows

Next up is Phase 2 which is building the AI assistant that teams can actually chat with to instantly pull insights from their own data.

It’s exciting to see how practical AI becomes when it’s built around real business needs.
If your company’s curious about using AI in a way that truly helps everyday work, this is exactly the kind of direction that makes a difference.

05/01/2026

In this video, I explained how the workflow I created for automated job search works. Feel free to check it out!

https://youtu.be/BP1wqDpikNk?si=e14Sv-0yuuJkT2U8

Feel free to send us a message if you want something the same.

In this video, I have created a workflow that will automatically scrapes jobs off Indeed. Feed it to OpenAI and scores the job for fit.