Aurora
Turn your product into an AI salesperson.
Point it at your website or docs. It builds a grounded knowledge base, then runs marketing, sales and solution agents - with a human approving everything that goes out.
Aurora
Point it at your website or docs. It builds a grounded knowledge base, then runs marketing, sales and solution agents - with a human approving everything that goes out.
Four containers, one command, your data stays on your machine.
One line on macOS, Linux or Windows. Or a Helm chart for Kubernetes.
Paste a website or docs URL. Aurora ingests it and builds a product profile.
SEO audits, content, outreach. You approve; then it publishes.
Everything below works in the PoC. Where something needs your credentials (publishing channels, keyword data), it says so.
Claims are checked against your own sources and cited. No invented pitches.
Audit any public page and score drafts. Keyword data is labelled measured or estimated.
Approval is tied to the exact content that goes out. Edit it and it needs approval again.
Research, outreach and proposals run on the server while you do something else.
Bring an OpenAI-compatible key, or run Ollama locally (GPU on Apple Silicon).
A signed, notarized SwiftUI client that connects to any Aurora server.
The fastest way to shape Aurora is a two-minute note: what worked, what hurt, what would make you use it for real.
Give feedback