Hey everyone. I've been trying to learn about voice AI for a side project, and I just went through the process of building a custom IVR voice with Resemble AI. Coming from a DevOps background, this was totally new territory for me!
I wanted to share my actual steps and what it ended up costing. I used their "Fill" method, which just needed a script and my own voice recording. The process was pretty straightforward—uploaded my audio, typed the script to match, and the platform trained the voice model. Total time from start to a usable voice was about 3 hours (mostly waiting for processing).
For costs, I stayed in the pay-as-you-go tier. Cloning the voice itself was free, which was great. I only paid for the generation. Making a 5-minute IVR message cost me roughly $4.80. The voice quality is surprisingly good, though sometimes it stumbles on technical words like "Kubernetes" 😅. Has anyone else here tried something similar? I'm curious how this might integrate into a CI/CD pipeline for app updates.