Andy O'Neil | AI Automation
Have an idea for technology to solve a problem but no team to build it? I bring business ideas to life with you. I'm Andy O'Neil.
- Indexed videos, last 90 days
- 12
- Latest publication
- Sep 28, 2026
- Audience
- ~5.4K subscribers
- Earliest in this view
- Jul 10, 2026
Latest videos
Why I Stopped Using AI for Every Make com Step (opens the original)
Read excerpt
Stop using AI first in Make.com. In this make.com tutorial I share the 3-rung ladder I use with clients: built-in make.com functions first, then code, then AI only when the input is truly unpredictable. You'll learn when to use AI in your workflows and how to stop quiet AI failures from corrupting production data. I'll show you how make.com functions handle most logic for free, when a short code step beats an AI call, and how to add make.com filters plus a make.com error handler so AI steps fail
I Got Paid $7 for 80 Minutes of Work (opens the original)
Read excerpt
Back in 2019, someone paid me $7 for an hour and twenty minutes of Integromat work. Integromat is now Make.com, and that call turned into how I build automations today: live, with you on the call. Co-building means you watch the build, answer questions as they come up, and tell me when I'm heading the wrong way. One step in your head can be six steps in a Make.com scenario, and the build is where those questions show up. Next video: what you actually own once your automation is built, where it l
I Built a Make.com Dashboard to Save My Clients and My Time (opens the original)
Read excerpt
Stop guessing if your automation is broken. I built a Make.com dashboard to give my clients instant visibility on failures. The biggest headache in business operations is not knowing whose fault it is when a process fails. Instead of managing panic and fielding unnecessary support calls, I created a central hub that tracks everything happening in the background. This approach goes beyond basic automation error handling. By implementing a dedicated Make.com dashboard, you can instantly see which
Your Make.com Automations Are Lying To You (Here's How To Check) (opens the original)
Read excerpt
Is your Make.com automation lying to you? Learn how to catch errors even when your scenarios run green. This guide is for owner-founders who need reliable systems. Green execution history in Make.com automation does not mean your data is correct. You might have a scenario that finishes without errors but still updates the wrong record or assigns incorrect invoice numbers. To fix this, I built a logic validation layer using AI. Instead of standard automation error handling, I use a two-module sys
I Built a Sports Tracker Chrome Extension (The Mistake I Almost Made) (opens the original)
Read excerpt
Building a custom Chrome extension with AI can be fast, but shipping reliable code requires a real review process. This isn't a step-by-step tutorial, but rather a look at my development workflow. I built a sports tracker called Pin the Game to handle live scores, but the real work wasn't the initial prompt. We cover how I structured the AI development to ensure the final product actually works as intended, rather than just spitting out a prototype. The biggest takeaway is what happened during m
Publishing over time
Last 90 days. Choose a month to open its work.
Recurring subjects
Named in the text we hold. One piece can cover several.
Audience
~5.4K subscribers
Measured Sep 19, 2026
Source's subscribers, not the number who saw an individual piece.
About this data
Counts cover the work we have indexed. Tone needs enough text and a confident classification. Excerpts and episode notes are not full articles or transcripts.
Identity or attribution wrong? Suggest a correction.