AgentSwarms Launches Self-Hosted Agentic AI & BI Platform with Full Data Control
Key Takeaways
- ▸Fully self-hostable, source-available platform using ELv2 license for agentic AI and business intelligence
- ▸Comprehensive feature set including multi-agent swarms, RAG knowledge bases, SQL query capabilities, BI dashboards, and document generation (PowerPoint, Word, Excel with live formulas)
- ▸Support for bring-your-own-keys from 10+ LLM providers (OpenRouter, OpenAI, Anthropic, Gemini, Bedrock, Azure, OCI, Qwen, Grok, Groq, Ollama, vLLM)
Summary
AgentSwarms has released a new self-hostable, source-available platform for building and deploying multi-agent AI systems and business intelligence solutions. Released under the ELv2 license, the platform enables organizations to create sophisticated agentic workflows while maintaining complete control over their data and infrastructure through bring-your-own-infrastructure deployment.
The platform includes a comprehensive feature set: a visual multi-agent swarm canvas for designing complex workflows, agent chat with visual BI capabilities, knowledge bases with RAG support, connections to major data warehouses (PostgreSQL, Snowflake, BigQuery, Redshift, etc.), BI dashboards with cross-filtering and PDF export, and a sandboxed Python notebook environment for development. All computation runs on user infrastructure, with Supabase as the backend and support for BYOK (bring-your-own-keys) from 10+ LLM providers including OpenAI, Anthropic, Gemini, Bedrock, and others.
Deployment is simplified to a single Docker command, with optional instance-wide LLM provider configuration for seamless user onboarding. The platform's design prioritizes developer control and transparency, showing the SQL queries and execution traces behind AI-generated answers and dashboards, enabling data verification rather than blind assertion.
- Integrations with major data sources including PostgreSQL, MySQL, Oracle, Snowflake, BigQuery, Redshift, Databricks, and S3-compatible storage
- Single Docker deployment with Supabase backend keeps all data and models on user's own infrastructure
Editorial Opinion
AgentSwarms represents a significant step toward truly decentralized AI infrastructure for enterprise use cases. By combining sophisticated multi-agent orchestration with comprehensive business intelligence capabilities in a self-hosted model, the platform addresses growing concerns about data sovereignty and vendor lock-in. The breadth of integrations—from LLM providers to data warehouses—and the inclusion of advanced developer tools like sandboxed notebooks suggest a maturity that could make this compelling for enterprises unwilling to trust cloud-based AI platforms with their proprietary data.
