Anyone using eleven labs for AI reception missed calls/call screening?

Hi, Question: I am a beginner to building AI agents. I have been researching building and offering AI Receptionists as my first entry for an offering to market to SMB's in person, face to face. I have spoken to a few decision makers who mentioned in their industry most calls are folks trying to sell them something. If they could use a AI receptionist to screen those it would be a valuable asset. With that in the background I am at a decision point: Learn Make/N8N and all that I'd have to to create custom agents. OR Use Go High Level A little background: I have trained new Pharma sales rep's on custom CRM platforms to insure compliance, I have been the go btwn Dev's and users on global custom reporting software, managed UAT's. I have trained offshore help desk managers and helped create knowledge bases for those implementations in colab with other team members. Please weigh in and give me your thoughts on which path you think is best and/or pitfalls with each.

I have been lurking in the background and am coming out lol. I hope to be more active as I learn how to create/build AI Voice Agent from scratch

To learn how to build inbound AI Voice agents for SMB tradespeople to capture missed calls, qualify, and gather contact info, how would you start?