Report Abuse

Basic Information

Helix is a private GenAI stack and control plane for building, deploying and running AI agents inside your own data center or VPC so organizations retain full data security and control. The repository provides the core control plane and tooling to define and deploy LLM apps (agents) using a helix.yaml declarative format and supports retrieval-augmented generation (RAG), API-calling, and vision capabilities. It includes an installer for Docker, helm charts for Kubernetes, and guidance to attach GPU runner nodes or point the control plane at OpenAI-compatible LLM APIs. The project also provides a local development guide, a web dashboard, documentation and community support for private deployments and an optional SaaS/paid launchpath for larger commercial use.

Links

App Details

Features
Private deployment of an agent control plane and attachable GPU runners, with an installer script for Docker and helm charts for Kubernetes. Declarative agent definitions via helix.yaml for building pipelines, knowledge/RAG integration, and API bindings. Vision support and API-calling built into agent workflows. A GPU scheduler that packs models into available GPU memory and dynamically loads and unloads models to optimize utilization. Ability to point the control plane at external OpenAI-compatible LLMs. Web dashboard defaulting to port 8080, developer local development guide, documentation, community Discord, and license terms for personal, educational, and small business use.
Use Cases
Helix helps organizations deploy secure, private GenAI agents without relying on external cloud services, enabling control over data residency and compliance. The declarative helix.yaml approach simplifies creating repeatable agent pipelines with RAG and API integrations, while the GPU scheduler and runner model improve resource efficiency for model hosting. Docker quickstart and Kubernetes helm charts make it practical to bootstrap test and production deployments. Integration with external OpenAI-compatible APIs provides flexibility to use hosted models or local GPU-backed models. Documentation, a local development guide and community channels ease onboarding. License terms and a commercial launch option clarify permitted usage and enterprise licensing paths.

Please fill the required fields*