I built a private AI job-intake assistant for a friend's electrical business

Описание и применение

Создан локальный AI‑ассистент для обработки заказов в электротехническом бизнесе, работающий офлайн и защищающий персональные данные клиентов.

Задача

Customers send messy requests via WhatsApp; the owner must manually parse and reply, leading to inefficiency and potential errors.

Результат

The assistant reduces manual effort, ensures consistent summaries, flags emergencies, and prevents fabricated pricing; all processing stays offline on the owner’s laptop.

Рабочий процесс

• 1. Customer submits request via public showcase form, which forwards to owner’s WhatsApp.
• 2. Owner pastes the WhatsApp message into the local intake script.
• 3. The script runs an open‑weight model (qwen2.5:1.5b or llama3.2:3b) via Ollama.
• 4. The model generates a one‑line summary, urgency flag, follow‑up questions, and a draft reply.
• 5. Owner reviews and sends the reply to the customer.
• 6. Public site displays job photos and contact details for future reference.

Доказательства

Customers message him all day with requests like rewiring damaged buildings, rewinding pump coils, faulty switches, flat wiring, solar and CCTV, burnt wires, urgent equipment repairs, and ATS and contactor prices.

Источник

I built a showcase site where a form sends each request to his WhatsApp, plus a private AI intake assistant on his laptop.

Источник

Runs an open-weight model locally with Ollama and turns messy customer messages into a clean job summary, urgency level, follow‑up questions and a draft reply.

Источник

Customer names, addresses and phone numbers stay on the owner's computer.

Источник

I tested 8 real customer messages four times and kept every run: v1 missed a fire risk: "the wire in my room has burnt" was marked NORMAL.

Источник