Am I Approaching This Business Wrong? · AI Automation Agency Hub

Am I Approaching This Business Wrong?

I’m a fresh software engineering graduate trying to start an AI automation agency, but I’m struggling to figure out what to sell.

Starting a startup has always been my dream. I want to build something of my own, get my first clients, and eventually turn it into a real business.

I decided to explore AI automation, targeting local businesses in the US, particularly HVAC companies.

But the more I research, the more confused I get.

Here’s what I’m struggling with:

Most HVAC businesses already use tools like ServiceTitan, Housecall Pro, Jobber, and others.

These platforms already offer scheduling, follow-ups, customer management, and many other automations.

So what am I supposed to sell them?

Lead follow-up automation?
Missed-call recovery?
AI receptionists?
Automated booking and customer communication?

These sound useful, but established software already handles many of these problems.

I’m struggling to find a genuine gap where I can provide real value instead of rebuilding something that already exists.

I’ve been researching businesses, looking for problems, and building demos to show potential clients. But sometimes, I feel like I’m building solutions that businesses don’t actually need.

And convincing a cold prospect to buy something they don't think they need feels almost impossible, especially through cold emails and DMs.

Here’s a little about my background:

I’m a fresh software engineering graduate. I can build full-stack applications using technologies like FastAPI, PostgreSQL,, LangGraph and more.

I use AI coding tools heavily, but I understand the code, architecture, and systems I’m building.

I’m not limited to basic no-code automations. I can build custom AI agents, backend systems, and complete applications if there’s a real problem worth solving.

So, I have a few questions for people who are already doing this:

Am I approaching the AI automation business model the wrong way?
Are basic automations actually selling well, or are people just hyping them up online?
How do you find problems that businesses will genuinely pay to solve?
Should I stick with AI automation, explore custom software development, or combine both?

I’m not looking for motivational advice or another list of automation ideas.

I want to hear from people who are actually getting clients, delivering solutions, and making money.

Tell me what works, what doesn't, and what you would do differently if you had to start again.

I genuinely want to build a real business. I just don't want to spend months building demos for problems nobody is willing to pay me to solve.

Any honest advice would mean a lot.

15 comments

Am I Approaching This Business Wrong?