We built Voxvol because every voice-AI platform we tried made us choose a provider, a prompt style, or a single-tenant model. We wanted all three to be a choice, not a constraint.
Voxvol started as an internal tool for running voice agents across a portfolio of client projects, each with different providers, compliance needs, and call volumes. Off-the-shelf platforms forced a single LLM, a single TTS voice, or a single-tenant model, none of which matched how agencies and platform teams actually operate.
Today Voxvol is built on an open-core voice engine, with a multi-tenant control plane for identity, provider policy, DID management, and quotas, so a single deployment can serve one team or forty client tenants with the same governance guarantees.
Voice AI infrastructure should be inspectable, portable, and yours to control, not a black box you rent by the minute from a single vendor.
We're hiring across engineering, and always happy to talk to design partners.