https://github.com/neural-maze/ava-whatsapp-agent-course
Meet Ava, the WhatsApp Agent
https://github.com/neural-maze/ava-whatsapp-agent-course
agent agent-based agentic-workflow agents stt tts vector-database
Last synced: 5 months ago
JSON representation
Meet Ava, the WhatsApp Agent
- Host: GitHub
- URL: https://github.com/neural-maze/ava-whatsapp-agent-course
- Owner: neural-maze
- License: mit
- Created: 2024-12-17T09:25:28.000Z (10 months ago)
- Default Branch: main
- Last Pushed: 2025-04-26T21:54:04.000Z (5 months ago)
- Last Synced: 2025-04-26T22:28:03.735Z (5 months ago)
- Topics: agent, agent-based, agentic-workflow, agents, stt, tts, vector-database
- Language: Python
- Homepage:
- Size: 25.5 MB
- Stars: 1,267
- Watchers: 20
- Forks: 292
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
- awesome-LangGraph - Ava WhatsApp Agent Course - term memory using LangGraph. | (📖 Additional Resources / Community Courses)
- awesome-repositories - neural-maze/ava-whatsapp-agent-course - Meet Ava, the WhatsApp Agent (Python)
README
📱 Ava 📱
Turning the Turing Test into a WhatsApp Agent
## Table of Contents
- [Table of Contents](#table-of-contents)
- [Course Overview](#course-overview)
- [Who is this course for?](#who-is-this-course-for)
- [What you'll get out of this course](#what-youll-get-out-of-this-course)
- [Getting started](#getting-started)
- [Course syllabus](#course-syllabus)
- [How much is this going to cost me?](#how-much-is-this-going-to-cost-me)
- [The tech stack](#the-tech-stack)
- [Contributors](#contributors)
- [License](#license)
## Course Overview
What happens when [two ML Engineers](#contributors) with a love for sci-fi movies team up? 🤔
You get **Ava**, a Whatsapp agent that can engage with users in a "realistic" way, inspired by the great film [Ex Machina](https://www.imdb.com/es-es/title/tt0470752/). Ok, you won't find a fully sentient robot here, but you **will** have some pretty interesting Whatsapp conversations.
By the end of this course, you'll have built your own Ava too, capable of:
* Receiving and sending Whatsapp messages 📲
* Understanding your voice 🗣️
* Recognizing your images 🖼️
* Sending voice notes back 🎤
* Sharing updates about its "daily activities" 🚣
* Sending you images of its current activities 🖼️
>You can think of it as a modern reinterpretation of the Turing Test 🤣
Excited? Let's get started!
---
📬 Stay Updated
Join The Neural Maze and learn to build AI Systems that actually work, from principles to production. Every Wednesday, directly to your inbox. Don't miss out!
🎥 Watch More Content
Join Jesús Copado on YouTube to explore how to build real AI projects—from voice agents to creative tools. Weekly videos with code, demos, and ideas that push what's possible with AI. Don't miss the next drop!
---
## Who is this course for?
This course is for Software Engineers, ML Engineers, and AI Engineers who want to level up by building complex end-to-end apps. It's not just a basic "Hello World" tutorial—it's a deep dive into making a production-ready WhatsApp agent.
## What you'll get out of this course
* Build a fully working WhatsApp agent you can chat with on your phone
* Get a solid understanding of how to build LangGraph workflows
* Set up a long-term memory system using Qdrant as a Vector Database
* Use Groq models to power AI Agent responses
* Implement STT systems using Whisper
* Implement TTS systems using ElevenLabs
* Generate high-quality images using diffusion models, like FLUX models
* Process images using VLM models, like llama-3.2-vision
* Create chat interfaces using Chainlit
* Deploy agentic applications to Cloud Run
* Connect agentic applications to the WhatsApp API
## Getting started
Before you begin the course, there are a few things you need to do.
I'm referring to the virtual environment creation, dependencies installation, `.env` file creation, etc. I know, it's very boring, but it's a necessary evil! 😅
All of this is detailed in the following doc: [GETTING STARTED.md](docs/GETTING_STARTED.md).
> Make sure you follow the instructions in the doc, as it's crucial for the course to work.
## Course syllabus
| Lesson Number | Written Lesson | Video Lesson | Description |
|---------------|----------------|--------------|-------------|
|
1 | [Project overview](https://theneuralmaze.substack.com/p/meet-ava-the-whatsapp-agent) | 
| Understand the project architecture and the tech stack. |
| 2 | [Dissecting Ava's brain](https://theneuralmaze.substack.com/p/dissecting-avas-brain) | 
| Learn the basics of LangGraph and implement complex workflows using this framework. |
| 3 | [Unlocking Ava's memories](https://theneuralmaze.substack.com/p/can-agents-get-nostalgic-about-the) | 
| Build a short-term memory system for graph state persistence and chat history. Also, implement a long-term memory system using Qdrant. |
| 4 | [Giving Ava a Voice](https://theneuralmaze.substack.com/p/the-ultimate-ai-voice-pipeline) | 
| Build a STT and a TTS pipeline to make Ava process input and output audio. |
| 5 | [Ava learns to see](https://theneuralmaze.substack.com/p/reading-images-drawing-dreams-vlms) | 
| Understand how to process images using VLM models. Implement an image generation pipeline using FLUX models. |
| 6 | [Ava installs Whatsapp](https://theneuralmaze.substack.com/p/connecting-an-ai-agent-to-whatsapp) | 
| Connect Ava to WhatsApp. Learn how to deploy a LangGraph application to Google Cloud Run. |
---
## How much is this going to cost me?
The awesome thing about this project is **you can run it on your own computer for free!**
The **free tiers** from Groq, ElevenLabs, Qdrant Cloud, and Together AI are more than enough to get you going.
If you want to try it out on Google Cloud Run, you can get a free account and get $300 in free credits. Even if you've already used up your free credits, Cloud Run is super cheap - so it will take just a buck or two for your experiments.
---
## The tech stack
Technology
Description
Powering the project with Llama 3.3, Llama 3.2 Vision, and Whisper. Groq models are awesome (and fast!!)
Serving as the long-term database, enabling our agent to recall details you shared months ago.
Deploying your containers easily to Google Cloud Platform
Learn how to build production-ready LangGraph workflows
Amazing TTS models
Behind Ava's image generation process
## Contributors
Miguel Otero Pedrido | Senior ML / AI Engineer
Founder of The Neural Maze. Rick and Morty fan.
LinkedIn
YouTube
The Neural Maze Newsletter
Jesús Copado | Senior ML / AI Engineer
Equal parts cinema fan and AI enthusiast.
YouTube
LinkedIn
## License
This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
---
📬 Stay Updated
Join The Neural Maze and learn to build AI Systems that actually work, from principles to production. Every Wednesday, directly to your inbox. Don't miss out!
🎥 Watch More Content
Join Jesús Copado on YouTube to explore how to build real AI projects—from voice agents to creative tools. Weekly videos with code, demos, and ideas that push what's possible with AI. Don't miss the next drop!