Case study · 2024 – 2026
Receptionist AI
No-code voice agents that call and book. No-code voice agents for outbound and inbound calls: qualify leads, book meetings, and fire SMS, email, WhatsApp, or calendar actions while the call is still live.
- Type
- saas
- Years
- 2024 – 2026
- Stack
- Node.js · Python · WebSockets · Telephony APIs · OpenAI API · ElevenLabs · PostgreSQL · Redis
The problem
Missed calls cost sales. Hiring callers around the clock is expensive. Most bots cannot finish a booking or hand off cleanly.
My role
Core engineer: telephony integration, lead automation, SaaS dashboard.
The hard part
Latency comes before model quality. STT, reasoning, and TTS must stay under ~500ms round-trip or the call feels dead. Call state, CRM rows, and calendar writes have to stay in sync when someone interrupts, switches language, or drops mid-booking. On top: flow builder, RAG knowledge bases, Zapier and webhooks.
Outcome
Live at tryreceptionist.ai. Inbound and outbound in 100+ countries, 11+ languages with auto-detect, sentiment scoring, drag-and-drop flows, and integrations from Salesforce to Cal.com. Free and Pro plans with a 14-day trial.
