My beautician can't answer the phone when her hands full, so I built her a Gemma 4 receptionist

Description and use

Aura, an AI receptionist powered by Gemma 4, answers salon clients 24/7, handles bookings, and forwards only owner‑specific questions, reducing workload and speeding up service.

Problem

A one‑person beauty salon owner cannot answer incoming client messages while treating clients, leading to a backlog of unanswered inquiries.

Outcome

Aura provides instant, accurate responses to client inquiries, reduces owner workload, and ensures timely handling of sensitive health or owner‑only questions. The system runs for about €15/month, while the owner can charge €19/month per salon, covering costs and generating profit.

Workflow

• 1. Client initiates chat on salon website or Telegram. 2. Message sent to Django API. 3. Code applies guardrails (off‑topic, health, owner‑only topics). 4. Gemma 4 returns structured JSON (reply + one action). 5. Python code validates action and performs it (book, cancel, handoff, none). 6. Owner receives a single push notification per client needing attention. 7. Owner replies via dashboard or Telegram, response sent back to client.

Evidence

Aura answers clients on the salon's website at any hour, in Spanish or English: prices, opening hours, recommendations, and booking, changing or cancelling appointments.

Source

When something needs Cris (an allergy, a complaint, a discount), Aura passes it on and Cris's phone buzzes, once per client and not once per message.

Source

Gemma never touches the database. For every message she returns structured JSON: a reply and one action.

Source

It costs almost nothing to run: $14 a month of hosting (the API and the Telegram bot) plus the domain, and the AI fits in Ollama Cloud's free tier.

Source