Send, receive, escalate.
Wherever your customers are.
Same AI, same knowledge base, same escalation rules — adapted to each channel's quirks (so you don't have to).
- Direct Meta integration — no Twilio middleman
- Read receipts + typing indicator
- 24-hour template fallback (compliant)
- Inbound media handled gracefully
- Paste a token — webhook configured automatically
- Per-org bots, multi-tenant from day one
- Customer + agent bots managed separately
- Trigger phrases for instant escalation
- For internal employees (role=user)
- Optimistic UI + Supabase realtime
- No bot tokens to wrangle
- Same agent dashboard, same escalation
Built on the four things
your competitors get wrong.
Confidence-based escalation, not a fallback button
Every inbound is scored for intent, frustration and lead-value. Cross your threshold? Hand off to a human — with full context. Your gate is configurable and auditable.
WhatsApp 24-hour rule, handled correctly
Most platforms break Meta's 24-hour session window. We auto-detect the boundary and send a pre-approved farewell template. No broken UX, no policy strikes.
One canonical pipeline across every channel
Telegram, WhatsApp, internal web — all normalize to the same shape and feed one workflow. Adding SMS or email later? It's an adapter, not a rewrite.
Multi-tenant RLS, open stack, self-hostable
Tenant isolation enforced at the Postgres layer (not just the API). The orchestration is a visual n8n graph — read it, audit it, modify it. No vendor lock-in.
Three steps. Under ten minutes.
No 90-day Zendesk onboarding. No solutions architect on a Zoom call.
Upload your knowledge
PDFs, spreadsheets, or markdown. We chunk, embed, and ground every reply in your real source-of-truth.
Connect your channels
Paste a Telegram bot token. Link your WhatsApp Cloud API number. Or just turn on internal web chat — no token needed.
Tune the AI, ship
Write its personality. Set escalation thresholds. Define trigger phrases. Watch it answer customers in real time.
See how we stack up.
| Capability | agent0 | Intercom Fin | Zendesk AI | Tidio | Chatbase |
|---|---|---|---|---|---|
| WhatsApp native (no markup) | $ extra | gated | shallow | ||
| Real human escalation | |||||
| Confidence-based gate | |||||
| 24h-window template fallback | |||||
| Multi-tenant out of the box | gated | ||||
| Self-hostable | |||||
| Editable orchestration | partial | ||||
| Predictable pricing | per-resolution | $115+/seat |
Comparison based on public documentation and G2 reviews as of 2026-04. Sources cited in our pricing FAQ.
The details that matter.
RAG over your KB
PDF, XLSX, MD → embedded → vector-searched. Replies cite sources, agents see similarity scores.
Human handoff
Classifier-driven escalation with full conversation context. Status flips AI → Human cleanly.
Real-time dashboard
Supabase realtime — agents see new messages instantly. No polling, no refresh button.
Multi-tenant RBAC
Platform admin, org admin, agent, and end-user roles. Dynamic role creation, email invite flow.
Editable workflow
The reply → escalate → send graph is a visual n8n workflow. Audit it, modify it, own it.
AI personality
Long-form markdown "AI Soul" you write yourself. Per-org tone, voice, escalation rules.
Predictable. Per-seat.
No per-resolution surprises.
14-day trial. No credit card. Cancel anything, anytime.
Starter
For solo founders + side projects
- 1 channel
- 500 conversations / mo
- 1 agent seat
- Email support
- Knowledge base (10 docs)
Pro
For growing support teams
- All 3 channels
- 5,000 conversations / mo
- 5 agent seats
- Multi-tenant orgs
- Priority support
- Unlimited KB documents
- WhatsApp templates
Enterprise
Self-host or private cloud
- Unlimited everything
- Self-host (docker compose)
- BYO LLM (Azure / Anthropic / local)
- SLA + dedicated support
- SOC 2 (in progress)
- On-prem deployment
Need EU data residency, BYO-LLM, or on-prem? Let's talk.
Frequently asked.
Ready to ship a support agent
that actually works?
Set up in under ten minutes. Cancel any time. No per-resolution accounting.