#118 · How to Start an AI Automation Agency ($0 Budget)

youtube ↗AI & Automation

What's your take on vibe coding — its future, and whether it could become a whole new business model? Which would you recommend between Bolt, Lovable, or Cursor? Also, which is the smarter long-term bet: offering automation services, or building actual software that automates things?

I think vibe coding could genuinely be the next big business model — you can build apps, software, or websites that solve niche problems with enough time invested. Between automation services and building software: I'd favor automation services, because building software means putting a lot of upfront capital and time into something in an era where the value of that knowledge is dropping fast — tools like Bolt, Lovable, and Cursor are increasingly capable of replacing exactly the kind of thing you'd have spent all that time building. On the tools themselves: I've used Bolt and like it (they even invited me to a boat party once, so take that with a grain of salt); I've used Lovable too, a cool product I covered in a video; and I've used Cursor for myself but haven't built anything major with it — it feels a bit different in nature from Bolt/Lovable, and it's hard to firmly recommend just one.

vibe codingboltlovablecursorbusiness model

Related answers

youtube ↗AI & Automation

How to transition to a SaaS model after building automations, including needed components like auth, storage, DB, billing, and scaling to 1k–10k users.

I think it's a good question, but I'm not a software engineer—I'm okay at it but not the most qualified. Still, I'll answer. First, I'd build a serverless backend using Make.com and Nadn (or similar), meaning you don't manage server infrastructure—you just send/receive events to webhook URLs. Once you have that automation, I'd build a front end with Bolt or Lovable (you get a free year of Bolt in Maker School). You'd vibe‑code the front end and connect it to the backend via webhooks. For authentication/storage/database, I'd use Superbase, which works out of the box with vibe‑coded tools like Bolt and Level. Then I'd add payments via Stripe, again trying to vibe‑code it with Lovable/Bolt, though I'd likely need help. Regarding scale, the hard truth is you shouldn't rely solely on vibe‑coded solutions past about 50 customers; they tend to have foundational security issues unless an actual engineer reviews them. I cited the Cluey hack as an example. So, get it off the ground quickly, but then invest in an engineer to secure customer data and prevent leaks—you have that responsibility when running a SaaS.

saasscalingsecurityengineer
youtube ↗AI & Automation

What is your workflow for vibe coding a SaaS app using AI tools like Replit, Cursor, and Cloud Code?

My vibe‑coding workflow starts with a voice transcription tool where I state three things: my objective (what the app should do), the context (my business, existing tools, clients), and the constraints (what I can or can't use, like needing N8N or avoiding certain platforms). Then I define what 'done' looks—for example, being able to send a request and get a nicely packaged JSON response within 500 ms. After that I just keep working, using the voice tool to feed the AI my goal, and verify my own work until the definition of done is met. Consistency matters: feeding the same structured prompt (objective, context, constraints, definition of done) yields reliable outputs, so I stick to that pattern rather than constantly changing how I prompt.

vibe codingai workflowprompt framework
youtube ↗AI & Automation

What top automation systems should a beginner focus on for high ROI, and is it better to offer only four or five automation services to a specific niche?

Great question. I actually just published a video that breaks this down, so I’ll point you there. The basic framework is a two‑column chart: the left side shows the types of automation you should avoid selling, and the right side lists the high‑ROI services you can sell. All of those right‑side items have been covered in videos on my channel over the past year, so you can stitch together a blueprint, add your own twists, and make sure it feels like your own offering rather than just a re‑branded template. Focus on a handful of high‑ROI automation services, customize them for your niche, and you’ll have a solid, marketable package.

automationsystemsroiniche
youtube ↗Niches

What niches are currently most in demand for automation in the US, such as HVAC, dentists, and law firms?

We have a bunch of people in Maker School experimenting with a wide landscape of niches, which is a benefit of being in the community. You can see which niches are working and which aren’t. HVAC, dentists, and law firms aren’t making a ton of money with automation because there aren’t enough of them to apply high‑volume, iterative strategies. I don’t recommend focusing on local services; instead go entirely digital and target industries with large headcounts that can benefit from automation without hurting profitability. The biggest boom right now is recruitment, which saw a surge of interest and will likely rise again. I also suggest experimenting with recruitment, older‑school agencies that haven’t caught up, MSPs, IT consultants, and similar businesses that can source leads automatically via tools like LinkedIn Sales Navigator.

automationnichesdigitalrecruitment