Building Applications with Vector Databases – Free & Open Source Notebooks

Welcome! 👋 This repository is a collection of hands-on notebooks originally inspired by DeepLearning.AI in collaboration with Pinecone.

✨ The twist? Unlike the original course notebooks that rely on OpenAI models (which cost money), this repo uses free and open-source alternatives — so you can run everything locally with zero extra cost!

---

🚀 What’s Inside?

• ✅ All course notebooks adapted for free use
• ✅ Replaced OpenAI models with Gemini
• ✅ Replaced OpenAI embeddings (1536-dim) with JinaAI embeddings (768-dim)
• ✅ Fixed dimension mismatches and other common errors
• ✅ No API billing worries — everything runs with free models

---

📖 Course Modules

This repo covers six powerful applications of vector databases and embeddings:

1. Semantic Search – Find documents based on meaning instead of exact keywords.
2. RAG (Retrieval-Augmented Generation) – Generate fact-based answers using retrieved knowledge + Gemini.
3. Recommender Systems – Suggest items to users using similarity and personalization.
4. Hybrid Search – Mix keyword + semantic search for more accurate results.
5. Facial Similarity Search – Compare and find visually similar faces with embeddings.
6. Anomaly Detection – Detect unusual patterns and outliers in data.

---

🛠️ Tech Stack

• Gemini → for text generation
• JinaAI → for embeddings
• Pinecone → for vector database
• Python + Jupyter Notebooks → for hands-on learning

---

📦 Installation

1. Clone this repo:

   git clone https://github.com/sdivyanshu90/Building-Applications-with-Vector-Databases.git
   cd Building-Applications-with-Vector-Databases

2. Create and activate a virtual environment (recommended):

   python -m venv venv
   source venv/bin/activate   # On Windows: venv\Scripts\activate

---

▶️ How to Use

1. Open Jupyter Lab or Jupyter Notebook:

   jupyter notebook

2. Explore the notebooks in order.

3. Experiment, tweak, and learn without worrying about API costs! 🎉

---

🎯 Learning Goals

By working through these notebooks, you’ll:

• Understand how to build with vector databases
• Learn how to generate and embed text without OpenAI
• Practice fixing real-world errors (dimension mismatches, etc.)
• Gain confidence working with free and open-source models

---

🙌 Acknowledgments

• DeepLearning.AI & Pinecone for the original course materials
• JinaAI for their free embedding models
• Open-source community for making this learning path possible 💙

---

💡 Who Is This Repo For?

• Students 🧑‍🎓
• Beginners in AI/ML 🤖
• Anyone who wants to learn without paying for APIs

---

⭐ Contribute

If you find bugs, fix errors, or have improvements, feel free to submit a pull request. Let’s learn and build together!