explaingit

nazrinzainal/syncera-ai-messenger

20TypeScriptAudience · pm founderComplexity · 3/5ActiveSetup · moderate

TLDR

Electron desktop app that mirrors a WhatsApp inbox via Baileys and adds AI reply suggestions through Ollama, LM Studio, or a free Pollinations cloud fallback.

Mindmap

mindmap
  root((syncera-ai-messenger))
    Inputs
      WhatsApp messages
      Local AI model
      Customer templates
    Outputs
      AI reply drafts
      Broadcasts
      Orders pipeline view
    Use Cases
      Run a small business WhatsApp inbox
      Auto draft replies with a local LLM
      Schedule reminders from chats
    Tech Stack
      TypeScript
      React
      Electron
      Baileys
      SQL.js

Things people build with this

USE CASE 1

Run a desktop WhatsApp inbox with AI suggested replies for a small shop

USE CASE 2

Send broadcasts and quick send messages to a contact list

USE CASE 3

Generate replies with a local Ollama model so customer data stays on the machine

USE CASE 4

Turn a WhatsApp chat into a calendar event from inside the app

Tech stack

TypeScriptReactElectronBaileysSQL.js

Getting it running

Difficulty · moderate Time to first run · 1h+

Local AI replies need Ollama or LM Studio running, and the WhatsApp pairing via Baileys is not officially supported by Meta so accounts risk a ban.

In plain English

Syncera is a desktop application that helps small businesses manage their WhatsApp conversations and add AI replies on top of them. It looks like the familiar WhatsApp inbox, with contacts, message threads, and conversation history, but it runs as its own program on your computer rather than in a browser. The project is built with React for the visible interface, Electron to wrap it as a desktop app, and a library called Baileys to talk to WhatsApp. Data is stored locally on your machine using a SQL.js database, which is a SQLite-style storage layer that lives in a file. The AI side lives in a file called electron/bridge/ai.js. It can talk to several different AI backends. It supports Ollama, which is a way to run language models on your own computer, as well as LM Studio, Jan AI, and LocalAI, which are similar local options. There is also a free cloud fallback called Pollinations AI, and a small helper that uses DuckDuckGo search to feed live web results into the AI's context. If you have Ollama installed, the app finds it on its default local address and uses whichever models you have pulled. Beyond chat, the app offers broadcasts, quick send, message templates, reminders, an orders view, a pipeline, and several reports and analytics screens. A short demo video is included that shows a WhatsApp chat being used to create a calendar event inside Syncera. The public repository deliberately leaves out private customer databases, WhatsApp authentication sessions, build outputs, installers, and unredacted screenshots. To run it, you install the npm dependencies and then start it in development mode. The project is by Nazrin Zainal from Malaysia, has 20 stars, and is written in TypeScript.

Copy-paste prompts

Prompt 1
Install dependencies for syncera-ai-messenger and start it in Electron development mode
Prompt 2
Configure the AI bridge to use Ollama on the default port with a llama3 model I already pulled
Prompt 3
Switch the backend from Ollama to Pollinations AI cloud fallback when no local model is running
Prompt 4
Walk me through pairing a WhatsApp account through Baileys without leaking the session to the public repo
Prompt 5
Add a new message template for order confirmations and wire it into the quick send view
Open on GitHub → Explain another repo

Generated 2026-05-22 · Model: sonnet-4-6 · Verify against the repo before relying on details.