FollowUp Local: a Gemma model on your laptop that tells you why the customer went quiet
Description and use
FollowUp Local uses a local Gemma 3 model to analyze chat logs from small sellers and draft follow‑up messages, keeping data private and enabling low‑cost deployment.
Problem
Small sellers lose deals when customers go quiet after saying they will check with someone, but they lack a way to understand the real concern and craft a timely follow‑up without exposing private chat logs to the cloud.
Outcome
In a CPU‑only Linux box, gemma3:4b processes a chat in ~19 s, correctly identifies stages such as waiting_on_someone_else, extracts relevant evidence, and drafts a follow‑up in the appropriate language. The tool runs entirely offline, preserving privacy and enabling a low‑cost one‑time purchase model.
Workflow
- Paste chat text into the UI or CLI
- Model runs locally via Ollama and Gemma 3
- Model outputs structured JSON with stage, real concern, evidence, and next message
- Seller reviews draft and sends manually
Evidence
FollowUp Local is a small tool for people who sell through chat: someone making candles and taking orders over Instagram DMs, a bike mechanic booking repairs on WhatsApp, a private tutor talking to parents on WeChat.
A real, unedited 32-second run on a CPU-only Linux box with no GPU. Paste a chat, click once, and gemma3:4b answers in about 19 seconds.
Nothing leaves your machine. No API key, no account, no per-message cost.
- Record date: 2026-10-18T05:56:16.416Z