How To Use
From trying it here to embedding it on your own site
Four steps, no account required for this demo.
- 1
Try it on this page first
Click the floating bubble in the bottom-right corner of any page on this site. Ask it something about NovaBuildz Chat — it's answering from a real (small) knowledge base, using a real model, not a canned script.
- 2
Pick a color and a default model
On the Home page, use the color swatches and model picker to preview how the widget looks and which AI answers your questions. The embed snippet below updates live to match your choices.
- 3
Copy the embed snippet
Copy the <script> tag shown on the Home page (or the generic version below) and paste it just before the closing </body> tag on your own site's HTML.
<script src="https://chat.novabuildz.com/widget.js" data-color="amber" data-position="bottom-right" data-name="Nova Assistant" defer ></script> - 4
That's it — no other setup
The script mounts itself into a Shadow DOM container, so it won't collide with your site's CSS or JavaScript in either direction. There's nothing else to install, configure, or initialize.
Script tag options
Every attribute is optional — omit any of them to use the default.
| Attribute | Values | What it does |
|---|---|---|
| data-color | amber (default) · blue · green · purple · slate | Accent color for the bubble, header, and send button. |
| data-position | bottom-right (default) · bottom-left | Which corner of the screen the widget anchors to. |
| data-name | any text | Display name shown in the chat header, e.g. "Acme Support". |
| data-provider | gemini · groq | Force a specific model instead of the backend's default (Gemini). |
How answers are generated
Every message is sent, along with recent conversation history, to your chosen AI model along with a small curated knowledge base about NovaBuildz Chat. The model is instructed to answer from that knowledge base when relevant, and to say it doesn't know rather than make something up otherwise. If the chosen provider errors — a bad key, an outage, a quota limit — the backend automatically retries with the next provider in line, so a visitor rarely sees a failure.