Fullview AI is your customizable AI support agent that goes far beyond the limits of a traditional chatbot. By combining Retrieval-Augmented Generation (RAG) with DOM analysis, Fullview AI can effectively “see” what’s happening on your customers’ screens. This allows it to guide users with contextual, on-screen assistance alongside conversational answers. It’s purpose-built for complex B2B software applications where functionality-related support dominates.
This guide will walk you through setup and first use so you can start deflecting tickets, accelerating resolutions, and giving customers an intelligent, frictionless support experience.
Set Up Your Fullview AI Workspace
Create an Account
Sign up with your company credentials.
Verify your email and log in to the Fullview dashboard.
Connect Your Knowledge Sources
For simplicity sake, we will begin by setting up the conversational AI chat agent. The latter part of this guide will focus on setting up the visual AI guidance agent.
Upload or link your documentation, help center, FAQs, and internal resources (this will also be required later on when we set up the visual AI guidance agent).
Fullview AI uses RAG to pull the right answers from your knowledge base, so the richer your data, the smarter your agent.
If your knowledge base is stored in Intercom, Zendesk or Salesforce you can connect them via the integrations.
You can also pick Scan Website if you prefer to just drop a link to your public help center articles or help center page for now (free plan is limited to 20 articles scanned). Alternatively, you can upload a file (CSV or JSON preferred) with your help center export.
Last but not least, you can also create snippets directly in Fullview's dashboard to create some quick instructions for the AI agent.
3. Begin testing!
Once you have added a knowledge source, this enables you to begin using the "playground mode" to test the conversational agent right away.
You can begin asking real questions in the chat interface to the right and review answers.
4. Personalize and invite your team
Want to show off your new AI support agent to your colleagues? Add customizations and share it with them by inviting them to your Fullview workspace.
Getting started with Fullview AI visual guidance.
Visual guidance requires Fullview's script to run in your application. This is required in order to read the DOM and generate the on-screen guidance.
Once that is done, you will be able to begin testing visual AI guidance in your own application using the Fullview chat widget. You can easily integrate Fullview with other chat widgets (e.g Zendesk Chat and others) - But we do recommend to save this step until you are ready to deploy Fullview AI in front of your real customers.
Testing Visual Guidance In Your Product
After the script has been installed, make sure to whitelist yourself and any other internal test users in your Fullview AI workspace. Whitelisted is necessary to bypass the publishing mechanism, as Fullview AI will only present visual guidance for topics that have been pre-approved by admins, so you can ensure quality of guidance and effectively use guardrails. Whitelisted allows you to bypass this for internal testing purposes.
Required: How to whitelist users.
Open the Fullview AI widget in your own application which will appear after you have installed the Fullview script above.
2. Test by asking a "how to" related questions where visual guidance would be helpful to your customers. Fullview AI will deliver a combo of text and offer the user "show me how to do it" call to action, which will trigger the visual guidance.
We recommend testing your top how-to related FAQs and "Publishing" the ones with highest out-of-the-box performance first.
What's next?
Once you are comfortable with the initial results, you can deploy Fullview AI to your customers. You can either continue to use the Fullview AI widget or integrate with your helpdesk software.
Remember you can always use AI Analytics to keep an eye on trends and topic performance, and use AI Replays to watch back autogenerated recordings of on-screen visual guidances in case, great for both QA and escalations!
Conclusion
With Fullview AI, you’re not just swapping in another chatbot - you’re upgrading to an intelligent, visual agent that actually understands your product and your customers’ context. Once set up, it will handle complex support queries, free up your team, and deliver faster resolutions.