🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

👯 Don't choose between fine-tuning & RAG – you can have both. If your RAG pipeline is struggling to understand the nuances in your data, you may want to try add a fine-tuned embedding model. What is embedding model fine-tuning? Fine-tuning an existing embedding model adapts

Weaviate 1.31 just dropped, bringing the search precision you've been asking for with new BM25 AND/OR operators! Here's what's new: 𝗕𝗠𝟮𝟱 𝗔𝗡𝗗/𝗢𝗥 𝗢𝗽𝗲𝗿𝗮𝘁𝗼𝗿𝘀 - Finally, granular keyword search control! • Need ALL terms to match? Use the AND operator (perfect for

Generic search is dead. Personalization is everything. After Query Agent and Transformation Agent, meet the Personalization Agent - search that adapts to each user. This isn't just another ranking tool. It's the first step toward truly personalized search that understands: 👤

Want to use multi-vector embeddings, but worried about your memory budget? 🧠💸 Multi-vector models like ColBERT and ColPali deliver incredible retrieval performance, but they come with a costly catch: larger memory footprints. A million documents can easily consume 40GB+ of

How StackAI is accelerating growth: Faced with pressure to integrate new AI features fast, Stack AI needed more than just basic vector search. They needed: - High-speed performance - Enterprise-grade security & compliance - Cost-effective scalability -A developer-first

Your favorite AI framework + Weaviate Query Agent = 🚀 Here are 𝟳 𝗱𝗶𝗳𝗳𝗲𝗿𝗲𝗻𝘁 𝘄𝗮𝘆𝘀 to integrate the Weaviate Query Agent into your existing AI stack. The Query Agent is a pre-built agentic service that answers natural language queries based on your Weaviate data -

Weaviate and NVIDIA are powering the next generation of AI apps with GPU-accelerated indexing and hybrid architectures built for scale 📈 NVIDIA’s AI factories are powering the future of scalable, intelligent infrastructure, and the new DGX B200 is built to meet that demand.

🎤 Thrilled to announce I'll be speaking at World Summit AI USA in San Francisco on June 18–19 💡 I’ll dive into all things Weaviate, with a special focus on our latest Weaviate Agents—how they work, what they can do, and why they matter for the future of autonomous AI-driven

What if your LLM could only say what you want it to say? Most people think “structured outputs” just mean getting valid JSON from a model. But it’s way deeper than that! On the Weaviate Podcast Ep. 119, Will Kurt and Even Better Cameron Pfiffer 📎 from .txt join Connor Shorten to reveal

Headed to the Databricks Data + AI Summit? Make sure to stop by booth D109 in the Artificial Intelligence Pavilion — that’s where the AI-native vector search magic happens. See how Weaviate powers production-grade RAG, semantic search, and AI apps with speed, scale, and

We are so excited to welcome Jolanta to our team today! 🎉 Jolanta Marczewsk* is joining as Product Designer and is based in Switzerland 🇨🇭 Welcome to Weaviate 💚

Sometimes the best architecture is the simplest one. Single-agent architectures are quietly powering some of the most effective AI applications today, transforming how we build RAG systems and modern AI solutions. What makes them so useful? 🎯 𝗟𝗼𝘄𝗲𝗿 𝗰𝗼𝗺𝗽𝗹𝗲𝘅𝗶𝘁𝘆 =

How does AI actually 𝑡ℎ𝑖𝑛𝑘 instead of just respond? The magic happens through agentic design patterns - the secret architectures that transform passive AI models into proactive problem-solvers. Here are just 4 of the most powerful patterns revolutionizing how AI works:

The best products aren't built in isolation – they're built WITH the people who use them. At #BerlinBuzzwords, Jessie de Groot and Marion Nehring will be diving into "From Culture to Open Source: Build Value-driven Communities" and sharing how community building transforms from

also if you missed my talk at #renderatl i have some fun online workshops on building/using agents in javascript, register!! - 📆 June 17: lu.ma/weaviate-query… - 🗓️ June 24: lu.ma/ws-agents-25-0…

We just hit 10,000 subscribers, and it’s all thanks to YOU! This milestone is a celebration of our incredible community. Your curiosity, feedback, and support mean the world to us! Thank you for helping us grow, learn, and innovate together 💚 If you're into vector search,

🙌 I'm preparing a talk (a good ol' demo!), and there's something in it that I'm—among many other things—extremely proud of: 𝗵𝗼𝘄 𝘄𝗲'𝗿𝗲 𝘀𝘁𝗮𝗿𝘁𝗶𝗻𝗴 𝘁𝗼 𝗯𝗹𝗲𝗻𝗱 𝗰𝗼𝗱𝗲 𝘄𝗶𝘁𝗵 𝗻𝗮𝘁𝘂𝗿𝗮𝗹 𝗹𝗮𝗻𝗴𝘂𝗮𝗴𝗲 𝘄𝗵𝗲𝗻 𝘄𝗼𝗿𝗸𝗶𝗻𝗴 𝘄𝗶𝘁𝗵 𝗪𝗲𝗮𝘃𝗶𝗮𝘁𝗲 🛡️

What happens when you give RAG the ability to think, plan, and use tools? You get 𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗥𝗔𝗚 - and it's changing how we build intelligent systems. Let’s break this down... 𝗪𝗵𝗮𝘁'𝘀 𝗮𝗻 𝗔𝗜 𝗔𝗴𝗲𝗻𝘁? Think of an agent as an AI system with some autonomy. Unlike

Your vector database is slowing down. Queries are taking longer. Memory is running out. Sound familiar? Let's talk about scaling Weaviate - because knowing 𝘄𝗵𝗲𝗻 and 𝗵𝗼𝘄 to scale makes the difference between graceful growth and system collapse. 𝗧𝘄𝗼 𝗣𝗮𝘁𝗵𝘀 𝘁𝗼

Looking for a cheaper alternative to agentic RAG? The solution might be late interaction. Late interaction is a retrieval technique where the similarity between a query and a document is calculated by comparing fine-grained, individual embedding representations (e.g., for each