These tutorials demonstrate real-world applications of Pixeltable through end-to-end workflows. Each tutorial includes complete code and detailed explanations.

All sample applications are open source and available on our Github.

​
Multimodal Processing

Audio Transcription Workflow

Build an end-to-end workflow for extracting, transcribing, and searching video audio using OpenAI Whisper and semantic indexing

Object Detection in Videos

Implement real-time object detection in videos with automatic frame extraction and processing

​
RAG Applications

Document Question Answering

Create a RAG system that summarizes PDFs and answers questions using ChatGPT with automatic incremental updates

Wikipedia Article Processing

Explore flexible RAG operations on Wikipedia articles with document splitting and multiple embedding types

​
Key Workflows

Audio Processing

Extract and transcribe audio from videos with automatic indexing

Computer Vision

Process video frames with ML models in real-time

Text Processing

Chunk and embed documents for semantic search

​
Implementation Features

Incremental Updates

Tables automatically update as new data arrives

Persistent Storage

Results remain available across sessions

Automated Workflow

Processing steps run automatically in sequence

​
Common Patterns

Media Processing Workflow

  1. Extract media (audio/frames)
  2. Process with ML models
  3. Index results for search

RAG Implementation

  1. Chunk documents
  2. Generate embeddings
  3. Build semantic index
  4. Query with LLMs

Each tutorial builds on Pixeltable fundamentals. New to Pixeltable? Start with our Pixeltable Basics tutorial or join our Discord community for help.

Was this page helpful?