LlamaIndex Unveils Cloud Service for Unstructured Data Agents
AI agents are all the rage these days, and while everyone's got their own take on what they are, the gist is they're AI tools that can do stuff on their own. The buzz around agents is pretty intense, but one company, LlamaIndex, was ahead of the curve. Started by ex-Uber research scientists Jerry Liu and Simon Suo back in 2023, LlamaIndex helps developers whip up custom agents that can work with unstructured data.
"LlamaIndex kicked off as a fun open-source project in November 2022," Liu shared with TechCrunch. "I got hooked on figuring out how large language models (LLMs) could tap into proprietary data not in their training set, so I built some tools to help developers index and use that data in their LLM apps."
With LlamaIndex's open-source software, which has been downloaded millions of times on GitHub, developers can craft custom agents that pull out info, whip up reports and insights, and even take specific actions. LlamaIndex offers data connectors and tools like LlamaParse, which turns messy data into something neat and usable for AI apps.
Sure, there are other open-source frameworks out there for building AI agents, but Liu says LlamaIndex stands out with its tools for data ingestion, management, indexing, and retrieval. It can hook up data from files like PDFs and PowerPoint slides, as well as apps like Notion and Slack, to an agent.
Big names like Salesforce, KPMG, and Carlyle are already using LlamaIndex, according to Liu.

A headshot of LlamaIndex co-founder Jerry Liu.Image Credits:LlamaIndex
"All these other solutions tackle specific issues at different layers of the generative AI stack, but then it's up to the developer to stitch them together to make a working agent," Liu explained. "That's a real headache that slows down getting agents into production. LlamaIndex's goal is to offer the most secure, accurate, and user-friendly platform for building complete knowledge agents."
LlamaIndex's next move is an enterprise service called LlamaCloud, built on their open-source stuff. It lets customers set up cloud-hosted agents that can handle and manipulate unstructured data in all sorts of formats.
You can roll out LlamaCloud as a software-as-a-service or in a virtual private cloud, and it comes with goodies like role-based access control and single sign-on, Liu mentioned.
To help fund LlamaCloud's growth, LlamaIndex just nabbed $19 million in a Series A round led by Norwest Venture Partners, with Greylock also chipping in. That brings their total funding to $27.5 million, and Liu says they'll use it to grow their 20-person team and keep developing their product.
"We've got enough cash to see us through the early stages of rolling out our platform commercially," Liu said. "We're betting big on developers playing a key role in bringing GenAI apps to businesses."
Related article
Vercel CEO Guillermo Rauch hints at IPO as AI agents boost revenue
Unlike many startups founded before ChatGPT that now struggle to find their footing in the AI era, Vercel, a decade-old development tool and website hosting platform, is thriving due to the surge of AI-generated applications and autonomous agents.“Wh
OpenAI Enhances Codex to Challenge Anthropic in Desktop AI Dominance
Currently, a quiet rivalry is unfolding between OpenAI and Anthropic over which company can deliver the most practical and powerful AI coding tools. So far, Anthropic appears to have the edge. As TechCrunch reported last week, Claude Code has become
AWS Neurosymbolic AI Delivers Verifiable Agent Automation for Regulated Sectors
AWS believes that making its Automated Reasoning Checks feature on Bedrock generally available will instill greater confidence in enterprises and regulated industries to adopt and deploy more AI applications and agents.The company further anticipates
Related Special Topic Recommendations
Comments (59)
0/500
Just read about LlamaIndex's new cloud service for unstructured data agents. Honestly, the agent hype is real, but seeing a company that was actually early to this is refreshing. Ex-Uber scientists building this? Makes you wonder what data challenges they faced there that sparked this idea. Curious about the pricing model though—hoping it's accessible for smaller devs and not just enterprise giants. The potential for automating messy data tasks is huge. 🤔
LlamaIndexのクラウドサービス発表か…非構造化データ向けエージェントって実際どのくらい使えるんだろう?🤔 前職でデータ整理に苦労した経験があるから、これが本当に機能すればかなり助かる人いるかも。でも元Uber研究者が起業ってところが気になるな、シリコンバレーではよくあるパターンだけど。
¡Menuda revolución con estos agentes de IA! Aunque muchos se suben ahora al carro, LlamaIndex ya vio el potencial hace tiempo. Me pregunto si su enfoque para datos no estructurados realmente marcará la diferencia frente a otras soluciones... 🤔 ¿Podrán mantener la ventaja técnica cuando todos lancen sus propias plataformas? Ojalá expliquen más sobre casos de uso reales.
This LlamaIndex cloud service sounds like a game-changer for handling messy data! 😎 I wonder how it stacks up against other AI agent tools in terms of speed and accuracy.
LlamaIndex का क्लाउड सर्विस कूल है लेकिन थोड़ा ओवरहाइप्ड है। अनस्ट्रक्चर्ड डेटा को हैंडल करने के लिए बढ़िया है, लेकिन सेटअप करना सिरदर्द था। अगर आप AI एजेंट्स में रुचि रखते हैं, तो इसे आजमाएं, लेकिन तकनीकी चुनौतियों के लिए तैयार रहें। 🤓💻
AI agents are all the rage these days, and while everyone's got their own take on what they are, the gist is they're AI tools that can do stuff on their own. The buzz around agents is pretty intense, but one company, LlamaIndex, was ahead of the curve. Started by ex-Uber research scientists Jerry Liu and Simon Suo back in 2023, LlamaIndex helps developers whip up custom agents that can work with unstructured data.
"LlamaIndex kicked off as a fun open-source project in November 2022," Liu shared with TechCrunch. "I got hooked on figuring out how large language models (LLMs) could tap into proprietary data not in their training set, so I built some tools to help developers index and use that data in their LLM apps."
With LlamaIndex's open-source software, which has been downloaded millions of times on GitHub, developers can craft custom agents that pull out info, whip up reports and insights, and even take specific actions. LlamaIndex offers data connectors and tools like LlamaParse, which turns messy data into something neat and usable for AI apps.
Sure, there are other open-source frameworks out there for building AI agents, but Liu says LlamaIndex stands out with its tools for data ingestion, management, indexing, and retrieval. It can hook up data from files like PDFs and PowerPoint slides, as well as apps like Notion and Slack, to an agent.
Big names like Salesforce, KPMG, and Carlyle are already using LlamaIndex, according to Liu.

"All these other solutions tackle specific issues at different layers of the generative AI stack, but then it's up to the developer to stitch them together to make a working agent," Liu explained. "That's a real headache that slows down getting agents into production. LlamaIndex's goal is to offer the most secure, accurate, and user-friendly platform for building complete knowledge agents."
LlamaIndex's next move is an enterprise service called LlamaCloud, built on their open-source stuff. It lets customers set up cloud-hosted agents that can handle and manipulate unstructured data in all sorts of formats.
You can roll out LlamaCloud as a software-as-a-service or in a virtual private cloud, and it comes with goodies like role-based access control and single sign-on, Liu mentioned.
To help fund LlamaCloud's growth, LlamaIndex just nabbed $19 million in a Series A round led by Norwest Venture Partners, with Greylock also chipping in. That brings their total funding to $27.5 million, and Liu says they'll use it to grow their 20-person team and keep developing their product.
"We've got enough cash to see us through the early stages of rolling out our platform commercially," Liu said. "We're betting big on developers playing a key role in bringing GenAI apps to businesses."
Vercel CEO Guillermo Rauch hints at IPO as AI agents boost revenue
Unlike many startups founded before ChatGPT that now struggle to find their footing in the AI era, Vercel, a decade-old development tool and website hosting platform, is thriving due to the surge of AI-generated applications and autonomous agents.“Wh
OpenAI Enhances Codex to Challenge Anthropic in Desktop AI Dominance
Currently, a quiet rivalry is unfolding between OpenAI and Anthropic over which company can deliver the most practical and powerful AI coding tools. So far, Anthropic appears to have the edge. As TechCrunch reported last week, Claude Code has become
AWS Neurosymbolic AI Delivers Verifiable Agent Automation for Regulated Sectors
AWS believes that making its Automated Reasoning Checks feature on Bedrock generally available will instill greater confidence in enterprises and regulated industries to adopt and deploy more AI applications and agents.The company further anticipates
Just read about LlamaIndex's new cloud service for unstructured data agents. Honestly, the agent hype is real, but seeing a company that was actually early to this is refreshing. Ex-Uber scientists building this? Makes you wonder what data challenges they faced there that sparked this idea. Curious about the pricing model though—hoping it's accessible for smaller devs and not just enterprise giants. The potential for automating messy data tasks is huge. 🤔
LlamaIndexのクラウドサービス発表か…非構造化データ向けエージェントって実際どのくらい使えるんだろう?🤔 前職でデータ整理に苦労した経験があるから、これが本当に機能すればかなり助かる人いるかも。でも元Uber研究者が起業ってところが気になるな、シリコンバレーではよくあるパターンだけど。
¡Menuda revolución con estos agentes de IA! Aunque muchos se suben ahora al carro, LlamaIndex ya vio el potencial hace tiempo. Me pregunto si su enfoque para datos no estructurados realmente marcará la diferencia frente a otras soluciones... 🤔 ¿Podrán mantener la ventaja técnica cuando todos lancen sus propias plataformas? Ojalá expliquen más sobre casos de uso reales.
This LlamaIndex cloud service sounds like a game-changer for handling messy data! 😎 I wonder how it stacks up against other AI agent tools in terms of speed and accuracy.
LlamaIndex का क्लाउड सर्विस कूल है लेकिन थोड़ा ओवरहाइप्ड है। अनस्ट्रक्चर्ड डेटा को हैंडल करने के लिए बढ़िया है, लेकिन सेटअप करना सिरदर्द था। अगर आप AI एजेंट्स में रुचि रखते हैं, तो इसे आजमाएं, लेकिन तकनीकी चुनौतियों के लिए तैयार रहें। 🤓💻





Home






