---
title: Video library — Product walkthroughs and demos ⬣ POMA AI
description: Guided walkthroughs of POMA AI's context-engineering stack.
canonical: https://www.poma-ai.com/video-library
generated: Markdown variant of the page above, built from the prerendered HTML
---
# 

Video library

Guided walkthroughs of POMA AI's context-engineering stack.

 

## Product

![Still frame from Fixing RAG's Biggest Problem — POMA AI Elevator Pitch](https://www.poma-ai.com/assets/video-elevator-pitch-qJu6GCh2.jpg)

1:14

### [Fixing RAG's Biggest Problem — POMA AI Elevator Pitch](https://www.poma-ai.com/video-library/fixing-rag-elevator-pitch)

POMA AI

Most RAG systems break documents into chunks without preserving structure or context. A one-minute look at that problem — and how POMA AI's structure-aware ingestion and chunking fixes it.

## Walkthroughs

![Still frame from POMA Grill Walkthrough: Connect Grill via MCP](https://www.poma-ai.com/assets/video-grill-walkthrough-D_cjMqLl.jpg)

2:47

### [POMA Grill Walkthrough: Connect Grill via MCP](https://www.poma-ai.com/video-library/grill-connect-via-mcp)

Ryan Bateman

End-to-end tour of the Grill loop: ingest documents in the POMA Console, watch them indexed into a project, then query that context from an agent over MCP.

![Still frame from POMA Grill Walkthrough: Connect Grill inside Claude Desktop](https://www.poma-ai.com/assets/video-grill-claude-desktop-BV1FRHrU.jpg)

1:36

### [POMA Grill Walkthrough: Connect Grill inside Claude Desktop](https://www.poma-ai.com/video-library/grill-connect-inside-claude-desktop)

Ryan Bateman

Add POMA Grill as a connector in Claude Desktop from the connector directory, then query your indexed documents directly inside a conversation.

## Webinars

![Still frame from POMA x Qdrant Webinar: how to use POMA AI's PrimeCut with your own Qdrant instance](https://www.poma-ai.com/assets/video-qdrant-webinar-C-oYHycF.jpg)

### [POMA x Qdrant Webinar: how to use POMA AI's PrimeCut with your own Qdrant instance](https://www.poma-ai.com/video-library/qdrant-primecut-webinar)

Dr. Alexander Kihm

Live walkthrough of pairing POMA AI's PrimeCut ingestion and chunking with your own Qdrant instance — from documents to structure-aware retrieval, with a runnable notebook to follow along.

[Open in Colab](https://colab.research.google.com/github/poma-ai/.github/blob/main/notebooks/qdrant/poma_meets_qdrant.ipynb) [View on GitHub](https://github.com/poma-ai/.github/blob/main/notebooks/qdrant/poma_meets_qdrant.ipynb)

The current notebook was updated and can slightly differ from the one in the video.