Vectorize Newsletter 2025-03-10
Welcome to the latest Vectorize newsletter! 🎉 We’re sharing new features, AI engineering insights, and useful resources to help you build smarter, more efficient AI applications.
✨ What’s New
Deep Research on your Private Data (Beta)
Why spend days piecing together reports when AI can do it for you? Deep Research lets you generate comprehensive, structured reports using your private data—enriched with the latest web information.
- Custom templates for tailored insights
- Automate research workflows with n8n
- Combined analysis of private and public information
Built-in Vectorize Database & Embedder
No setup required—Vectorize now includes a built-in vector database and embedder! You can still bring your own, but if you just want to get started quickly, you’re all set.
The New Visual RAG Pipeline Editor
We’ve redesigned the RAG Pipeline Editor for a cleaner, more intuitive experience. Now you can build and deploy pipelines faster than ever.
📖 Check out our walkthrough and give it a whirl!
Extraction Tester
Not sure how your documents will process? Test different extraction methods before setting up a pipeline. Choose between our fast extractor or Vectorize Iris for more advanced document processing.
Vectorize API + Client (Beta)
Manage your pipelines, connectors, AI platforms, and vector databases programmatically with the Vectorize API—now in beta. We’ve also released a new client library to make it even easier to get started!
- Python and Node.js client libraries
- Full control of pipelines, connectors, and vector DBs
- Programmatic file uploads for ingestion
Vectorize Iris
Got messy PDFs with tables, diagrams, or complex layouts? Vectorize Iris is our fine-tuned vision model that extracts and chunks text from PDFs in a single step—no extra setup needed.
💡AI Engineering Insights
Building a Context-Sensitive AI Assistant
We built an AI assistant that provides context-aware help in Vectorize, using data from our docs, Discord, and Intercom. It even adjusts its responses based on what users are doing in the app!
💡 Want to build your own AI assistant? After creating a RAG pipeline, download a ready-to-use Next.js chatbot pre-configured with your retrieval endpoint!
📚 Resource Corner
Check out our latest videos!
🚀 Want better RAG performance?
Your vector database is only as good as your retrieval strategy. Learn how to test different vectorization approaches and choose the best one—without the guesswork.
🚀 RAG Pipelines That Actually Work
RAG pipelines can ingest data from a variety of sources—but how do you optimize extraction, chunking, and retrieval settings for the best results? This video walks you through building and fine-tuning a pipeline that actually delivers relevant responses.
🚀 Build a Slack AI Assistant That Actually Knows Things
Want a Slack bot that answers real questions, not just generic AI-generated responses? Learn how to connect n8n, Vectorize, and Slack to create a fully functional AI assistant powered by your own documents and data.
⚡Coming Up
Google Drive OAuth is almost here! Simplify your source connection process with quick, secure authentication—no more service accounts needed.
📬 Stay Connected
Got questions? We’re here to help!
💬 Join us on Discord
📧 Email us at contact@vectorize.io
Happy building! 🚀
The Vectorize Team