Vigyata.AI
Is this your channel?

Qdrant & Make: Set up a RAG vector database in minutes

184 views· 4 likes· 23:32· Jan 19, 2026

🛍️ Products Mentioned (16)

If this video helps you, subscribe to the channel https://www.youtube.com/@pickertgmbh?sub_confirmation=1 Quickly set up a reliable Qdrant vector database for RAG and populate it via Make using the API — get clean embeddings, filterable payloads, and reliable queries. What you'll take away: - Configure a Qdrant cluster and create a collection (Simple Single Embedding). - Understand chunks and generate them from your data. - Use OpenAI Text-Embedding-3-Small to generate embeddings. - Add payload fields to enable filtered searches. - Generate UUIDs for unique IDs. - Perform batch uploads of multiple chunks via the API. - Query the data with and without filters. Keywords: Qdrant, vector database, Make integration, RAG, embeddings, chunking ➡️CHAPTERS 00:00:00 RAG systems and vector databases: basics and use cases 00:00:32 Why Qdrant: free, EU hosting and advantages over Pinecone 00:02:06 Qdrant account, cluster setup and API key management 00:03:32 Creating collections: types, Global Search and collection settings 00:04:34 Configuring embeddings: dimensions, metric and Text-Embedding-3-Small 00:06:11 Preparing the data workflow: content, variables and example collection 00:08:40 Creating chunks: chunk size, overlap and sensible splitting 00:12:46 Generating embeddings and building JSON payloads for Qdrant 00:14:17 Generating UUIDs and batch uploading vectors to Qdrant 00:19:34 Querying in Qdrant: vector search, filters, limit and score evaluation This is the AI-translated version of our YouTube-Video originally posted on our German YT-Channel @pickertgmbh 👉 SOFTWARE, TOOLS & DEALS + MAKE.com*: https://www.make.com/en/register?pc=pickertgmbh + Airtable*: https://airtable.com/invite/r/GbOmyMYx + OpenAI API: https://platform.openai.com/ + Elevenlabs*:https://try.elevenlabs.io/7uf5u0hvrmzs + HeyGen*: https://heygen.com/?sid=rewardful&via=sven-o + meinGPT*: https://partner.meingpt.com/pickert-gmb-h + 0CodeKit*: https://my.0codekit.com/en/auth/register?via=sven-o + Fillout*: https://www.fillout.com?ref=pickert + tl;dv*: https://tldv.cello.so/cbsANc1a33V *Affiliate links 👉 30-MINUTE STRATEGY CALL You want to get started but don’t know how? Sven O. Rimmelspacher has been working in quality and process optimization for over 30 years. Let’s talk – book your free call with Sven here: https://link.pickert.gmbh/termin-sor 👉 FOLLOW US Follow us, give us a like, and subscribe to our channels! 💻 Our website: https://www.pickert.de 🖊️ Our blog: https://www.pickert.de/blog ➡️ Our LinkedIn page: https://www.linkedin.com/company/pickertgmbh 👉 WHY WORK WITH US? At Pickert, we’ve been focused on quality and process optimization for over 40 years. We offer solutions for process automation and AI – and our team can implement exactly what you need, whether it’s a challenge or a full-service package. As part of the family-owned corporate group about ZERO GmbH, we’re part of a network of companies dedicated to more quality in every area of life. about ZERO GmbH: https://www.about-zero.de Quality Miners GmbH: https://www.quality-miners.de Rocket Routine GmbH: https://www.rocket-routine.com Factory Excellence Network GmbH: https://www.factory-excellence.com Pickert GmbH | Creating space to focus. Enabling clarity. Driving innovation.

About This Video

In diesem Video zeige ich dir, wie du in wenigen Minuten eine saubere RAG-Basis aufbaust: Qdrant als Vector Database und Make.com als Automations-Schicht für Uploads und Queries per API. Ich gehe zuerst die Basics durch (wofür RAG-Systeme da sind, was eine Vector Database eigentlich speichert) und erkläre, warum ich Qdrant hier gerne nutze: kostenlos startbar, EU-Hosting und insgesamt ein sehr pragmatisches Setup – gerade im Vergleich zu Pinecone. Danach setze ich mit dir gemeinsam einen Qdrant-Cluster auf, lege eine Collection im „Simple Single Embedding“-Setup an und konfiguriere die Embedding-Parameter (Dimensionen, Metrik) passend zu OpenAI Text-Embedding-3-Small. Der wichtigste Teil ist dann der Daten-Workflow: Ich zeige dir, wie ich Content in sinnvolle Chunks splitte (Chunk Size + Overlap), daraus Embeddings generiere und ein sauberes JSON baue – inklusive Payload-Feldern, damit du später wirklich filtern kannst. Zum Schluss mache ich Batch-Uploads mit UUIDs als IDs und demonstriere die Suche: einmal pure Vector Search und einmal mit Filtern, inklusive Limit und Score-Bewertung, damit du die Ergebnisse zuverlässig interpretieren kannst.

Frequently Asked Questions

🎬 More from Intelligent Process Automation