While companies today increasingly recognize the potential of custom AI agents, many still struggle to build and scale these applications. An Economist Impact report shows that only 37% of executives believe their GenAI applications are production-ready, a figure that drops to just 29% among practitioners. A major barrier is the absence of a robust agent workflow framework that can efficiently ingest, index, and query large volumes of enterprise data. This challenge is compounded by the fact that most enterprise data remains unstructured, making it especially difficult to parse, analyze, and extract actionable insights.
As more organizations turn to Databricks Mosaic AI to develop AI agent applications, we continue to offer innovations that enable our customers to build robust agent workflows and unlock insights from their unstructured data. That’s why we are thrilled to announce Databricks Ventures’ investment in LlamaIndex, a leading platform that enables enterprises to build production-ready knowledge agents capable of retrieving information, synthesizing insights, and taking actions on the most complex enterprise data. Customers today are already leveraging LlamaIndex’s integration with MLFlow and Mosaic AI Vector Search to build knowledge agents on the Databricks Data Intelligence Platform. This investment will further strengthen our relationship with LlamaIndex and pave the way for deeper technical collaborations.
Why LlamaIndex
Much like Databricks, LlamaIndex began as a developer-focused open-source project and quickly gained traction – surpassing 3 million monthly downloads and earning over 40,000 GitHub stars in just 2 years. LlamaIndex enables companies to build and deploy custom knowledge AI agents in days through its Workflows framework. This accelerated timeline reduces engineering overhead and helps enterprises achieve data intelligence faster. Additionally, LlamaIndex’s robust parsing and extraction capabilities accurately convert complex, unstructured data into formats optimized for LLMs, greatly enhancing the quality of AI agents. That’s why forward-thinking organizations such as Carlyle and Cemex use LlamaIndex to power their AI agent initiatives. Databricks Ventures’ investment in LlamaIndex will further accelerate its ability to serve the more than 60,000 customers currently on the waitlist for its enterprise offerings.
What’s next for LlamaIndex and Databricks
Following our investment, we look forward to collaborating more closely with LlamaIndex to deepen the integration between our platforms, making it easier for Databricks users to leverage LlamaParse (unstructured data parsing) and LlamaExtract (structured data extraction) directly on the Databricks Data Intelligence Platform.
Come see us both at the Data + AI Summit in San Francisco in June. Stay tuned for future updates as Databricks and LlamaIndex continue to collaborate to simplify AI agent deployment, helping organizations innovate faster and achieve greater ROI on GenAI!