wavefront

Wavefront is an open-source enterprise middleware for building AI-powered workflows and agents. It acts as a unified API layer and orchestration platform that connects various data sources, knowledge bases, and AI models into cohesive pipelines. You can configure multiple agents and workflows to run across enterprise data sources (databases, OLAP/OLTP, cloud storage, APIs) with built-in authentication, RBAC, observability, and guardrails. The server exposes a consistent interface for triggering tasks, routing results, and managing agent orchestration, enabling production-grade AI applications with enterprise controls.

To get started, install and run the Wavefront server, then use the provided tooling to define agents, connect data sources, and compose workflows. The platform supports integration with multiple LLM/SLM providers, retrieval-augmented generation through internal MCP connectors, and modular AI application composition. Use the CLI (when available) or the programmatic API to create agents, configure workspaces, and monitor execution through Grafana/Prometheus telemetry dashboards.

Prerequisites:

• Node.js (v16+ recommended) and npm
• Git

1. Clone the repository

git clone https://github.com/rootflo/wavefront.git
cd wavefront

2. Install dependencies for the wavefront-server

npm install --prefix wavefront-server

3. Configure environment (example)

# Create a config file or set environment variables as needed
export WAVEFRONT_ENV=production
export WAVEFRONT_CONFIG=./config/default.json

4. Run the Wavefront server

node wavefront-server/index.js

5. Verify the server is running by hitting the API endpoint or opening the dashboard as documented in the Wavefront docs.

Notes and tips:

• This server acts as middleware; you’ll typically connect data sources, models, and agents via MCP connectors.
• Ensure proper RBAC configuration for production use. Plan to integrate with your existing Google/Entra SSO if needed.
• Telemetry is available via Grafana/Prometheus; configure dashboards to monitor agent performance and guardrails.
• If you upgrade, APIs may change; refer to ROADMAP.md and the platform docs for latest integration details.
• If you encounter hostname or firewall issues, verify network access between the Wavefront server and your data sources.
• Environment variables and config paths are configurable; use a centralized config.json for reproducible deployments.