Getting Started

OpenAI Assistant is a ready-to-deploy chatbot application built on the OpenAI Assistants API. It wraps your existing OpenAI Assistant (configured on platform.openai.com) with a chat UI that you can embed in your website or use standalone. Available as an open web service in Eyevinn Open Source Cloud.

Prerequisites

Step 1: Create an OpenAI Assistant (if you don't have one)

  1. Go to platform.openai.com/assistants
  2. Click Create assistant
  3. Give it a name, instructions, and choose a model (e.g. gpt-4o)
  4. Copy the Assistant ID — it looks like asst_abc123...

Step 2: Store your OpenAI API key as a secret

Navigate to the OpenAI Assistant service in the Eyevinn OSC web console. Go to the Service Secrets tab and click New Secret.

Create a secret named openaikey with your OpenAI API key (starts with sk-...).

Step 3: Create an OpenAI Assistant instance

Go to the My openai-assistants tab and click Create openai-assistant. Fill in:

Field Description
Name Name for your instance (alphanumeric and underscores only)
OpenAiApiKey {{secrets.openaikey}}
AssistantId Your OpenAI Assistant ID, e.g. asst_abc123...
AppUrl (optional) URL of the website where you will embed the chat widget

Click the instance card when the status is green and "running" to open the chatbot UI.

Step 4: Embed in your website (optional)

If you provided an AppUrl, the service configures CORS to allow embedding from that domain. Add an iframe to your page:

<iframe
  src="https://your-instance-url"
  width="400"
  height="600"
  style="border: none;"
></iframe>

CLI usage

osc create boldare-openai-assistant my-chatbot \
  -o OpenAiApiKey="sk-..." \
  -o AssistantId="asst_abc123..."

Resources