Skip to content
Module 7Deployment Surfaces & Adoption Strategy

One Agent, Every Surface

You can share an agent's instructions, knowledge base, and configured skills across multiple business applications. Availability, access, presentation, and conversation continuity depend on each surface's configuration; test each enabled surface.

The Five Deployment Surfaces

1. Your Website (raia Live Chat Widget)

What it is: An embeddable web component that provides a branded chat experience on any webpage.

How to deploy:

  1. In the Launch Pad, go to the Launch tab
  2. Select View Embed Code and copy the JavaScript embed snippet

Worked example

The examples use fictional data. Follow the same checks with approved, non-sensitive information for your own task.

Open the agent, then Launch. Agent Status shows Public, Private or Inactive; Private means only organization members can chat with this agent. View Embed Code gives the snippet for your site. Select Test Agent to try a sandbox conversation. If no chat appears, keep the agent Private and report the issue; switching to Public broadens access and is not a preview fix. Changing the status changes who can reach the agent, so check with your admin first.

Launch tab for a Private agent.
  1. Highlight 1: Launch tab
  2. Highlight 2: View Embed Code
  3. Highlight 3: Agent Status
  1. Paste it into your website's HTML (before the closing </body> tag)
  2. Open the page and check that the widget appears and answers a test question

Worked example

The examples use fictional data. Follow the same checks with approved, non-sensitive information for your own task.

Ask a realistic question with a known answer, then compare the reply with your source. Also ask for a fact the source does not contain: the agent should acknowledge the gap rather than invent it. Review enabled actions first; a test conversation can still trigger them.

The public test agent's reply on the local test page.
  1. Highlight 1: Test question
  2. Highlight 2: Agent reply

After End Chat, the widget offers How did we do? The visitor can choose a rating and enter a comment before Submit Feedback. Use that feedback to identify conversations worth reviewing; a rating does not change the agent’s instructions by itself.

Feedback survey after End Chat, before a rating is chosen.
  1. Highlight 1: How did we do?
  2. Highlight 2: Rating choices
  3. Highlight 3: Submit Feedback

For the full setup, see the Live Chat skill guide.

Customization options:

  • Brand colors, logo, and avatar
  • Welcome message and quick-reply buttons
  • Pre-chat lead capture forms (name, email, phone)
  • Position (bottom-right, bottom-left)
  • Security Key and access restrictions (documented controls to configure, then test on your own site)

Advanced: Use the SDK Studio at https://sdk.raia.run to build fully custom chat interfaces with the raia JavaScript SDK.

2. raia Copilot (Internal Teams)

What it is: The unified omnichannel inbox where your internal team interacts with agents directly.

Use cases:

  • Internal knowledge queries ("What's our refund policy for enterprise clients?")
  • Testing and debugging before deployment
  • Monitoring and taking over customer conversations
  • Providing feedback for review, correction, and re-testing; feedback does not automatically retrain or permanently fix the agent

3. raia Browser Extension

What it is: A Chrome/Edge extension that gives employees access to raia agents from any webpage they visit.

How it works:

  • Install the extension from the Chrome Web Store and connect an agent with its API key
  • Open the chat drawer from the raia icon on a webpage
  • With Read page content turned on, the extension sends page text to the agent as context, so you can ask about the page you're viewing
  • See the browser extension guide for setup and settings

Ideal for: Sales teams researching prospects, support teams working in external tools, anyone who needs agent assistance without switching tabs.

4. Microsoft Teams / Slack

What it is: A raia app in your team's collaboration platform, connected to an agent through that platform's skill.

How to deploy:

  • Add the Microsoft Teams Skill or the Slack skill to the agent, then choose which channels and people may use it
  • Follow the Microsoft Teams guide or the Slack guide
  • Team members reach the agent only where you have allowed it, such as an authorized channel or a direct message
  • The agent uses the same instructions and knowledge; test it in each place you enable

5. SaaS Applications (Zendesk, Salesforce, Zoho, etc.)

What it is: Agent intelligence embedded directly into existing helpdesk, CRM, or business tools.

Example patterns (designs you build and test, not built-in features):

  • Zendesk: raia agent auto-drafts ticket responses; human agents review and send
  • Salesforce: Agent provides real-time coaching to sales reps during calls
  • Zoho: Agent handles first-response on incoming support tickets

These integrations typically use the raia API or webhook-based workflows via n8n to pass data between systems.

Driving Adoption Across Teams

Deploying the technology is only half the battle. Driving adoption requires a deliberate strategy:

  1. Start with one use case: Pick the highest-pain, lowest-risk scenario (e.g., answering repetitive IT questions)
  2. Measure results: Track resolution rate, response time, and user satisfaction for 2–4 weeks
  3. Share wins: Present concrete metrics to stakeholders (for example, the share of Tier 1 tickets resolved without human help)
  4. Expand incrementally: Add new channels, new departments, new use cases one at a time

Key Principle: Instructions and knowledge are shared across surfaces, but each channel has its own skill settings and access controls. Configure and test each surface you enable.