LlamaIndex
Get started
Talk to us

© 2024 LlamaIndex

LlamaIndex • Sep 24, 2024

LlamaIndex Newsletter 2024-09-24

Greetings, Llama Fans! 🦙

Welcome to this week's LlamaIndex newsletter! We're excited to share LlamaParse Premium, which enhances your LLM applications with advanced parsing capabilities, our new Multimodal RAG feature for efficient handling of product manuals in LlamaCloud, and check out RAGApp v0.1 for creating multi-agent applications effortlessly. Don't miss the latest tutorials from our community!

If you haven't explored LlamaCloud yet, make sure to sign up and get in touch with us to discuss your specific enterprise use case.

🤩 The highlights:

  • LlamaParse Premium Launched: Upgrade your LLM applications with LlamaParse Premium, featuring advanced document parsing that merges multimodal models with robust text extraction, improving RAG accuracy and adding capabilities like diagram-to-markdown translation and LaTeX equation outputs. Blogpost, Tweet.
  • Multimodal RAG for Product Manuals: New feature in LlamaCloud enables rapid building of RAG systems that understand complex product manuals, streamlining multimodal knowledge assistant creation. Blogpost, Tweet.
  • Story Generation Agent Guide: A new guide on developing a dynamic 'choose-your-own-adventure' story agent that incorporates human feedback to refine story outcomes and enhance narrative precision. Docs, Tweet.
  • RAGApp v0.1 Launch: Introducing RAGApp v0.1, which simplifies the creation of multi-agent applications. Users can now add agents like researchers and analysts to generate news articles through an interactive chat interface that provides streamed answers with sources. Code, Tweet.
  • Llama Researcher Project: Rohan's Llama Researcher project excels in multi-step, agentic report generation, managing tasks, researching, and report synthesis efficiently through LlamaIndex workflows. Code, Tweet.

🗺️ LlamaCloud And LlamaParse:

  • We have launched LlamaParse Premium to upgrade your LLM applications with our advanced document parser that merges multimodal models with robust text extraction, enhancing RAG accuracy and offering features like diagram to markdown translation, LaTeX equation outputs, and improved content detection. Blogpost, Tweet.
  • We have launched a Multimodal RAG feature for Product Manuals: Quickly build RAG systems with LlamaCloud that understand complex product manuals. LlamaCloud streamlines the setup of indexing, search, and retrieval pipelines, allowing users to create comprehensive multimodal knowledge assistants in minutes instead of weeks. Blogpost, Tweet.

✨ Framework:

  1. We've launched RAGApp v0.1, enabling users to easily create multi-agent applications without coding. Add agents like researchers and analysts to generate news articles via an interactive chat interface that streams answers with sources. Code, Tweet.
  2. Guide to Building a Story Generation Agent with Human-in-the-Loop to develop a dynamic 'choose-your-own-adventure' story agent that integrates human feedback to refine story outcomes and improve accuracy. Docs, Tweet.
  3. We (Zac) have integrated NUDGE, a non-parametric embedding fine-tuning approach, into LlamaIndex, enabling rapid optimization of data embeddings directly with minimal compute and no need for original model details. Notebook, Tweet.
  4. We've partnered and integrated with Opik by Cometml to improve RAG and agent applications, offering automated logging, evaluation, and integration features for development and production. Docs, Tweet.

💡 Use-case:

  • Llama Researcher project by Rohan, excels in multi-step, agentic report generation by inputting tasks, researching sub-topics with TavilyAI, and synthesizing final reports, all streamlined through LlamaIndex workflows. Code, Tweet.

✍️ Community:

  • Benito Martin’s tutorial on Creating a RAG application with AWS CDK as IaC, Qdrant and LlamaIndex.
  • NVIDIA’s video tutorial on Building LLM Assistants with LlamaIndex, NVIDIA NIM, and Milvus.
  • Akshay’s tutorial on deploying a private Llama 3.1 RAG API using LitServe and LlamaIndex.
  • Karan Vaidya’s tutorial on building AI Product Manager to Automate User Feedback into Action Points with Human-in-the-Loop using Composio and LlamaIndex.
  • Pavan Nagula’s tutorial on User-Centric RAG with LlamaIndex Multi-Agent System and Qdrant.

👥 Recruitment: