Topic 5 of 16 in the learning path
Embeddings & Vector Databases
Embedding models, similarity search, vector stores and indexing.
Reels and posts (15)
- Google AI Edge Foresight: On-Device Meeting Notes with EmbeddingGemma 2 · Melvin Vivas, X: Google AI Edge Foresight is an on-device macOS meeting companion.
- Vector Database vs. Graph Database: When to Use Each (and Both for RAG) · Bashiri Smith, Facebook · 0:50: This short reel compares vector databases and graph databases.
- Embeddings: what they mean for businesses · Melvin Vivas, X: The post shares an AIBackends blog article about business use cases for embeddings.
- Try EmbeddingGemma 2 multimodal embeddings in Colab · Melvin Vivas, X: EmbeddingGemma 2 is Google's multimodal embedding model.
- EmbeddingGemma 2 Colab Notebook for FAQ Semantic Search · Melvin Vivas, X: Melvin uploaded a Google Colab notebook that shows how to use EmbeddingGemma 2.
- EmbeddingGemma 2 released on Hugging Face · Melvin Vivas, X: The creator announces that Google's EmbeddingGemma 2 embeddings model is now on Hugging Face.
- EmbeddingGemma 2: One On-Device Model for Multimodal Embeddings · Melvin Vivas, X · 0:53: Melvin Vivas shares Google's announcement of EmbeddingGemma 2.
- EmbeddingGemma 2: open multimodal embeddings model for on-device use · Melvin Vivas, X: The creator shares Google's announcement of EmbeddingGemma 2, an open embeddings model built to run on-device, and says Gemma releases excite him more than Gemini ones.
- Liquid AI LFM 2.5 Encoder: a CPU-friendly encoder model · Melvin Vivas, X: Points out Liquid AI's LFM 2.5 Encoder, released in July 2026, which runs on CPU.
- How RAG Finds the Right Document Fast: Graph-Based Vector Search · Bashiri Smith, Facebook · 1:08: A skit about an interview question: how does a RAG app find the right document for a question?
- Sentence Transformers v6: ColBERT-Style Multi-Vector Models Fully Supported · Melvin Vivas, X: Sentence Transformers v6.0 has been released.
- Building Interactive AI Concept Demos with Claude Design (Embeddings Example) · Melvin Vivas, X · 0:19: In this 19-second X post with no narration, Melvin Vivas shows an interactive demo of how embeddings work that he made with Claude Design, Anthropic's tool for making visual design
- Gemini Embedding 2: One Vector Space for Five Modalities · Melvin Vivas, X: Gemini Embedding 2 is now generally available.
- Fine-Tuning MiniLM Embeddings with Synthetic Data, Running on CPU · Melvin Vivas, X: A practical case study in text matching.
- Fine-tuning all-MiniLM-L6-v2 to Beat OpenAI Embeddings · Melvin Vivas, X: The creator trained a custom embedding model starting from the small open-source all-MiniLM-L6-v2.
Watch, free (4)
- Word Embedding and Word2Vec, Clearly Explained!!! (StatQuest with Josh Starmer) · video · youtube.com · free
StatQuest's visual explanation of word embeddings and how Word2Vec learns them.
Mentioned in: 6 Free Videos to Move from Software Engineer to AI Engineer (Bashiri Smith on Facebook · notes), How to Relearn LLMs & RAG in 2026: A 7-Step Roadmap with Free Resources (Bashiri Smith on Facebook · notes) - Understanding and Applying Text Embeddings (DeepLearning.AI) · course · deeplearning.ai · free
A DeepLearning.AI course on embeddings and semantic search. Free with a DeepLearning.AI account during its platform beta; certificates are paid.
Mentioned in: How to Relearn LLMs & RAG in 2026: A 7-Step Roadmap with Free Resources (Bashiri Smith on Facebook · notes) - Vector Databases: from Embeddings to Applications (DeepLearning.AI) · course · deeplearning.ai · free
A DeepLearning.AI short course on how vector databases work and how to build applications with them. Free with a DeepLearning.AI account during its platform beta; certificates are paid.
Mentioned in: How to Relearn LLMs & RAG in 2026: A 7-Step Roadmap with Free Resources (Bashiri Smith on Facebook · notes) - What is a Vector Database? Powering Semantic Search & AI Applications (IBM Technology) · video · youtube.com · free
An IBM video introducing vector databases.
Mentioned in: How to Relearn LLMs & RAG in 2026: A 7-Step Roadmap with Free Resources (Bashiri Smith on Facebook · notes)
Read and use (16)
- EmbeddingGemma 2 · tool · huggingface.co · free
Google's open 740M-parameter multimodal embedding model that puts text, images, video and audio into one shared vector space for on-device use.
Mentioned in: Google AI Edge Foresight: On-Device Meeting Notes with EmbeddingGemma 2 (Melvin Vivas on X · notes), Try EmbeddingGemma 2 multimodal embeddings in Colab (Melvin Vivas on X · notes), A day of work done from a phone with Grok Bot agents (Melvin Vivas on X · notes), EmbeddingGemma 2 Colab Notebook for FAQ Semantic Search (Melvin Vivas on X · notes) and 3 more - Pinecone · tool · pinecone.io · free
A managed, production vector database (free tier).
Mentioned in: Basic RAG Pipeline in 60 Seconds: From Documents to Grounded Answers (Bashiri Smith on Facebook · notes), Step-by-Step Roadmap to a $200K+ AI Engineering Role (Bashiri Smith on Facebook · notes), How RAG Works: Chunking, Embedding, Vector Storage, and Retrieval (Bashiri Smith on Facebook · notes), How RAG Works Under the Hood: Chunking, Embedding, Storage, Retrieval (Bashiri Smith on Facebook · notes) and 1 more
In a shared PDF: The AI Pivot Field Guide - shared in this reel on Facebook, this reel on Facebook - Weaviate · tool · weaviate.io · free
An open-source vector database for storing embeddings and running semantic and hybrid search.
Mentioned in: Basic RAG Pipeline in 60 Seconds: From Documents to Grounded Answers (Bashiri Smith on Facebook · notes), Step-by-Step Roadmap to a $200K+ AI Engineering Role (Bashiri Smith on Facebook · notes), How RAG Works: Chunking, Embedding, Vector Storage, and Retrieval (Bashiri Smith on Facebook · notes), How RAG Works Under the Hood: Chunking, Embedding, Storage, Retrieval (Bashiri Smith on Facebook · notes) and 1 more - pgvector · tool · github.com · free
An open-source Postgres extension for storing embeddings and running vector similarity search.
Mentioned in: Step-by-Step Roadmap to a $200K+ AI Engineering Role (Bashiri Smith on Facebook · notes), How RAG Works: Chunking, Embedding, Vector Storage, and Retrieval (Bashiri Smith on Facebook · notes), 7 Habits to Become an AI Engineer: Books, Tooling, Research & Shipping (Bashiri Smith on Facebook · notes), How RAG Works Under the Hood: Chunking, Embedding, Storage, Retrieval (Bashiri Smith on Facebook · notes) - all-MiniLM-L6-v2 · tool · huggingface.co · free
A small open-source sentence-transformers embedding model that maps text to 384-dimensional vectors.
Mentioned in: Fine-Tuning MiniLM Embeddings with Synthetic Data, Running on CPU (Melvin Vivas on X · notes), Fine-tuning all-MiniLM-L6-v2 to Beat OpenAI Embeddings (Melvin Vivas on X · notes) - Chroma · tool · trychroma.com · free
The easiest open-source vector database to start with.
Mentioned in: The 5 Levels of RAG: From Naive Retrieval to Agentic RAG (Bashiri Smith on Facebook · notes)
In a shared PDF: The AI Pivot Field Guide - shared in this reel on Facebook, this reel on Facebook - OpenAI text-embedding-3-small · tool · platform.openai.com · paid
OpenAI's small paid embedding model, available through its API.
Mentioned in: Fine-Tuning MiniLM Embeddings with Synthetic Data, Running on CPU (Melvin Vivas on X · notes), Fine-tuning all-MiniLM-L6-v2 to Beat OpenAI Embeddings (Melvin Vivas on X · notes) - Sentence Transformers · tool · sbert.net · free
Open-source Python library for training and running embedding, sparse, reranker and (from v6) multi-vector models.
Mentioned in: Try EmbeddingGemma 2 multimodal embeddings in Colab (Melvin Vivas on X · notes), Sentence Transformers v6: ColBERT-Style Multi-Vector Models Fully Supported (Melvin Vivas on X · notes) - aibackends (Python package) · tool · pypi.org · free
A Python package installed with pip and shown in the post's caption; it seems to be connected to the demo.
Mentioned in: Building Interactive AI Concept Demos with Claude Design (Embeddings Example) (Melvin Vivas on X · notes) - EmbeddingGemma 2 Developer Guide · docs · developers.googleblog.com · free
Google's guide to embedding multimodal content with EmbeddingGemma 2 using sentence-transformers and other frameworks.
Mentioned in: Try EmbeddingGemma 2 multimodal embeddings in Colab (Melvin Vivas on X · notes) - Embeddings business use case (AIBackends blog) · article · aibackends.com · free
Blog article on what embeddings mean for businesses and how they can be used.
Mentioned in: Embeddings: what they mean for businesses (Melvin Vivas on X · notes) - FAISS · tool · faiss.ai · free
Fast local similarity search from Meta.
In a shared PDF: The AI Pivot Field Guide - shared in this reel on Facebook, this reel on Facebook - Gemini Embedding 2 · tool · blog.google · check price · open in a browser to verify
Google's multimodal embedding model, with one vector space for text, image, video, audio and PDF.
Mentioned in: Gemini Embedding 2: One Vector Space for Five Modalities (Melvin Vivas on X · notes) - Gemma Cookbook · repo · github.com · free
Google's collection of example notebooks for building with Gemma models, including EmbeddingGemma.
Mentioned in: EmbeddingGemma 2: One On-Device Model for Multimodal Embeddings (Melvin Vivas on X · notes) - LFM 2.5 Encoder · tool · liquid.ai · free
Liquid AI's encoder model, released in July 2026, that can run on CPU.
Mentioned in: Liquid AI LFM 2.5 Encoder: a CPU-friendly encoder model (Melvin Vivas on X · notes) - OpenAI Ada embeddings · tool · platform.openai.com · paid
OpenAI embedding model used by llama_index by default.
Mentioned in: Q&A Over Your Own PDFs with LlamaIndex, OpenAI and Python (2023) (Melvin Vivas on melvinvivas.com · notes)
Build
- Build a local semantic search tool that embeds your notes, docs, and transcripts with an on-device embedding model such as EmbeddingGemma. (from Google AI Edge Foresight: On-Device Meeting Notes with EmbeddingGemma 2)
- Build a RAG system that uses vector search to retrieve relevant content and a graph database to follow the relationships between the retrieved items. (from Vector Database vs. Graph Database: When to Use Each (and Both for RAG))
- Build a search that works across text, code, images, video and audio using EmbeddingGemma 2 (from Try EmbeddingGemma 2 multimodal embeddings in Colab)
- Build a semantic search over your own FAQ with EmbeddingGemma 2 embeddings and cosine ranking. (from EmbeddingGemma 2 Colab Notebook for FAQ Semantic Search)
- Build cross-modal search where a text query finds matching images, video or audio, all using one embedding model. (from EmbeddingGemma 2: One On-Device Model for Multimodal Embeddings)
- Build on-device classification or clustering of mixed media (text, images, audio) with EmbeddingGemma 2. (from EmbeddingGemma 2: One On-Device Model for Multimodal Embeddings)
- Build an on-device semantic search over mixed media (text, images, audio) with EmbeddingGemma 2, using Matryoshka-truncated embeddings to save storage. (from EmbeddingGemma 2: open multimodal embeddings model for on-device use)
- Compare dense, sparse and ColBERT-style multi-vector retrieval on your own RAG dataset. (from Sentence Transformers v6: ColBERT-Style Multi-Vector Models Fully Supported)
- Use Claude Design to build an interactive visual explainer of how embeddings work: text becomes vectors, and similar meanings sit close together. (from Building Interactive AI Concept Demos with Claude Design (Embeddings Example))
- Build a multimodal search/RAG system that searches text, images, audio and PDFs in one Gemini Embedding 2 index. (from Gemini Embedding 2: One Vector Space for Five Modalities)
- Fine-tune all-MiniLM-L6-v2 on synthetic pairs for a domain-specific text-matching task, publish it to Hugging Face, and serve it from a CPU-only Docker container (from Fine-Tuning MiniLM Embeddings with Synthetic Data, Running on CPU)
- Fine-tune all-MiniLM-L6-v2 on domain-specific pairs and compare retrieval quality against OpenAI text-embedding-3-small. (from Fine-tuning all-MiniLM-L6-v2 to Beat OpenAI Embeddings)