Qdrant

🚀 Missed the live session? Evgeniya Sukhodolskaya and Max Tkacz demonstrated how to design and deploy production-ready AI agents with n8n’s low-code AI starter kit and Qdrant’s vector database. 👉 Watch the full recording to learn how to scale and deploy AI agents that solve real-world problems: https://lnkd.in/dP8AZXnR

📖 𝐀 𝐏𝐫𝐚𝐜𝐭𝐢𝐜𝐚𝐥 𝐆𝐮𝐢𝐝𝐞 𝐭𝐨 𝐁𝐮𝐢𝐥𝐝𝐢𝐧𝐠 𝐂𝐨𝐥𝐥𝐚𝐛𝐨𝐫𝐚𝐭𝐢𝐯𝐞 𝐀𝐈 𝐀𝐠𝐞𝐧𝐭 𝐓𝐞𝐚𝐦𝐬 Kameshwara Pavan Kumar Mantha wrote a manual on creating an agentic system for news search using AskNews, Qdrant, and phidata. 𝘛𝘩𝘪𝘯𝘬 𝘰𝘧 𝘢𝘨𝘦𝘯𝘵𝘪𝘤 𝘴𝘺𝘴𝘵𝘦𝘮𝘴 𝘢𝘴 𝘵𝘦𝘢𝘮𝘴 𝘸𝘩𝘦𝘳𝘦 𝘦𝘷𝘦𝘳𝘺 𝘱𝘢𝘳𝘵𝘪𝘤𝘪𝘱𝘢𝘯𝘵 𝘩𝘢𝘴 𝘢 𝘤𝘭𝘦𝘢𝘳𝘭𝘺 𝘥𝘦𝘧𝘪𝘯𝘦𝘥 𝘳𝘰𝘭𝘦 (𝘢𝘯𝘥 𝘺𝘰𝘶’𝘳𝘦 𝘵𝘩𝘦 𝘰𝘯𝘦 𝘥𝘦𝘧𝘪𝘯𝘪𝘯𝘨 𝘪𝘵). This guide walks you through: ✅ Crafting prompts for AI agents: a step-by-step approach to writing precise instructions for Agents. ✅ Setting up AI Agents tools. ✅ Managing the memory of AI agents. 👉 https://lnkd.in/df3aMJQE

Working with async Python apps? Make your Qdrant integration non-blocking! Perfect for FastAPI and other async frameworks 🔄 ✨ Maintains async flow & event loop responsiveness 📖 Learn how: https://lnkd.in/dd3aFN3C

⚖️ 𝐁𝐮𝐢𝐥𝐝 𝐚𝐧 𝐀𝐈 𝐋𝐞𝐠𝐚𝐥 𝐓𝐞𝐚𝐦 𝐫𝐮𝐧 𝐛𝐲 𝐀𝐈 𝐀𝐠𝐞𝐧𝐭𝐬 The second in a row tutorial from Unwind AI, kudos to their explanation style—clear and concise! This tutorial demonstrates how to configure a team of 𝐀𝐈 𝐀𝐠𝐞𝐧𝐭 𝐥𝐚𝐰𝐲𝐞𝐫𝐬 using GPT-4o, phidata, and Qdrant. The resulting Streamlit application, based on an input PDF document, can perform contract review, legal research, risk assessment, compliance checks and even work with custom queries. The tutorial is adaptable to many other complex domains, so we encourage you to try! ✍ By: Shubham Saboo & Gargi Gupta 📖 Tutorial: https://lnkd.in/dQC_ehcY 💻 Codebase: https://lnkd.in/drHxzJnd 📹 Demo: https://lnkd.in/dsh2xsNJ

🚀 High-Performance RAG Agent with Cohere ⌘R Building a production-grade RAG pipeline doesn’t have to be complicated. In this tutorial from Unwind AI, learn to create a fast, scalable system using Cohere’s Command R7B, Qdrant, LangChain, and LangGraph. Command R7B redefines enterprise RAG with its 128k context window and native in-line citations, delivering a rare combination of speed, precision, and efficiency. 💡 What You’ll Build: ➡️ A system that processes PDFs, embeds them with Cohere, and stores embeddings in Qdrant for fast retrieval. ➡️ A retrieval engine with a fallback to web search for unmatched queries. ➡️ Workflow orchestration with LangGraph to tie it all together. ✍ By: Shubham Saboo & Gargi Gupta 📖 Tutorial: https://lnkd.in/dXghw5qC 💻 Codebase: https://lnkd.in/gwZJEf7J 📹 Demo: https://lnkd.in/dKuv_7R3

Last week, the Qdrant team flew in from all corners of the world for our annual retreat in Gran Canaria. 🌍✈️ We had incredible workshops and brainstorming sessions, but the real magic happened in late-night conversations that sparked new ideas, honest and unfiltered discussions, and moments like watching a teammate’s goofy dance moves and remembering this is the same person who crushed a major product launch last quarter. We laughed, learned, stargazed through telescopes, and exchanged bits of each other’s languages. Here’s to another year of solving tough problems, celebrating small wins, and occasionally embarrassing ourselves in front of each other. 🌌✨ To the team that makes it all worthwhile: thank you. ❤️ And if you’ve ever wondered what it’s like to work with a group like this, we’re always looking for more incredible people to join us. (PS: Good magic tricks skills optional, but highly encouraged.) 

🔍 𝐀𝐈-𝐏𝐨𝐰𝐞𝐫𝐞𝐝 𝐖𝐞𝐛 𝐒𝐞𝐚𝐫𝐜𝐡, 𝐏𝐫𝐢𝐯𝐚𝐭𝐞 𝐚𝐧𝐝 𝐒𝐞𝐥𝐟-𝐇𝐨𝐬𝐭𝐞𝐝 PrAIvateSearch is a self-hosted, open-source search tool that delivers smarter, contextual answers. It runs entirely on your machine, ensuring full privacy and control. 1️⃣ Images are locally captioned with Florence-v2-large to generate search queries. 2️⃣ Results are fetched, processed, and indexed entirely on your machine. 3️⃣ Responses are generated using Qwen-2.5-3B, enriched with RAG with Qdrant for deeper context. 4️⃣ Add local datasets or custom knowledge bases for even more control. Built by Astra Clelia Bertelli 🚀 Read the full implementation blog 👉 https://lnkd.in/dHAb-wFt Try PrAIvateSearch 👉 https://lnkd.in/dD77FpRq

Voiceflow enables anyone to create complex AI agents without the need for managing infrastructure. To ensure their no-code platform remains scalable and reliable, they chose Qdrant for its seamless multi-node scaling, precise metadata-driven search, and secure, SOC2-compliant infrastructure. "The Qdrant team is very responsive, ships features quickly, and is a great partner to build with." – Denys Linkov, Machine Learning Lead at Voiceflow 👉 Read the full story: https://lnkd.in/dSMBfawV

Quick preview of the next release feature. It is Qdrant running in cluster mode on 𝐒𝐭𝐞𝐚𝐦 𝐃𝐞𝐜𝐤 😁 and leveraging 𝐆𝐏𝐔 𝐚𝐜𝐜𝐞𝐥𝐞𝐫𝐚𝐭𝐢𝐨𝐧 to construct the HNSW ANN index. Just a few seconds for 100K data points index building. Coming soon. 🚀

Vector similarity alone isn't always enough - real applications need to combine semantic search with traditional filtering. Let's explore how Qdrant's powerful filtering language helps you build complex queries that match your exact needs! 🔍 💡 Remember you can use dot notation for nested fields and even filter array elements with the [] syntax! 📚 Full filtering guide: https://lnkd.in/damY8Srs