CoSupport AI
What is Agent Cloud?
Agent Cloud provides a platform for creating data-connected AI agents that utilize Retrieval-Augmented Generation (RAG) with access to current vector data. It features a built-in data pipeline capable of handling information from over 260 data sources, alongside uploaded files like PDF, DOCX, TXT, and CSV. This pipeline manages the splitting, chunking, and embedding processes necessary to prepare data for vector storage.
Designed with flexibility and security in mind, Agent Cloud is LLM agnostic, supporting connections to open-source models, locally hosted models for enhanced privacy, or cloud services like OpenAI and Azure OpenAI. It integrates with major vector databases such as Qdrant and Pinecone (for self-hosted instances) and allows users to control data synchronization frequency to ensure agents access up-to-date information. The platform emphasizes secure data access, enabling teams within an organization to utilize AI applications safely.
Features
- Data Synchronization: Connect and sync data from over 260 sources and various file types.
- RAG Pipeline: End-to-end processing including splitting, chunking, embedding, and vector storage.
- LLM Agnostic: Supports OpenAI, Azure OpenAI, open-source, and local LLMs.
- Vector DB Integration: Compatible with major vector databases like Qdrant and Pinecone.
- Data Control: Select specific tables, columns, or fields for syncing and embedding.
- Embedding Flexibility: Use cloud embedding models or local options via fastembed.
- Configurable Sync Frequency: Set data updates manually, on a schedule, or via cron expression.
- Secure Team Access: Manage user permissions for secure AI application usage.
- Open Source Option: Provides a community edition with an AGPL 3.0 license.
Use Cases
- Building RAG chat applications with access to current data.
- Creating AI agents connected to diverse internal data sources.
- Deploying secure AI applications within enterprise environments.
- Developing specialized AI tools for Sales, HR, IT, and Customer Support teams.
- Enabling private and secure chat interactions with organizational data.
- Automating data ingestion and preparation for AI models.
FAQs
-
What is the software license for the open-source version?
The open-source Community edition uses the AGPL 3.0 license. -
Can I use a local Large Language Model with Agent Cloud?
Yes, Agent Cloud supports any local LLM with an OpenAI compatible endpoint, such as those run via LM Studio or Ollama. -
Which cloud LLM models do you support?
Currently, Agent Cloud supports OpenAI and Azure OpenAI, with plans to support more cloud providers. -
What file upload formats are supported?
Agent Cloud supports PDF, DOCX, TXT, CSV, and XLSX file formats for local uploads. -
How can data retrieval be truly private?
For true privacy, it is recommended to deploy the open-source version of Agent Cloud on your own infrastructure and potentially use a locally hosted LLM.
Related Queries
Helpful for people in the following professions
Featured Tools
Join Our Newsletter
Stay updated with the latest AI tools, news, and offers by subscribing to our weekly newsletter.
