Self-Hosted AI Chatbot

A practical deployment review for teams with stricter infrastructure requirements

A self-hosted AI chatbot can make sense when your network, data residency, provider, or infrastructure policies rule out a standard SaaS deployment. Use this guide to decide what must be owned and operated before launch.

When self-hosting makes sense

Consider a self-hosted deployment when data must remain in a specific jurisdiction, private network, or customer-controlled environment and a hosted service cannot satisfy that boundary.

  • Jurisdiction need
  • Private network
  • Provider control

What your team must operate

Before choosing self-hosting, assign owners for infrastructure, upgrades, monitoring, backups, secrets, AI provider access, incident response, and customer support.

  • Ops owner
  • Secrets
  • Upgrade process

What self-hosting does not guarantee

Running software in your environment does not automatically create a compliance certification, managed uptime SLA, or global residency program. Those require separate evidence and operating controls.

  • No certification claim
  • No SLA claim
  • No residency badge

Our self-hosted product: DChat

ZChat itself is cloud-hosted. RichScripts also builds DChat (dchat.com), a separate self-hosted live chat product you install on your own servers. It is sold as a subscription: Community edition free with unlimited agents, Premium $19 and Enterprise $49 per agent per month. It is a different product from ZChat cloud, so check its own feature list rather than assuming the two match.

  • Community: free, unlimited agents
  • Premium: $19 per agent/month
  • Enterprise: $49 per agent/month

Self-hosted decisions should be recorded before launch so support, security, and infrastructure teams share the same assumptions.