Most people don't want "AI infrastructure." They want outcomes, and the Fabric AI Assistant just got significantly better at delivering them.
They want fewer support tickets. They want customers to find answers faster. They want to add an AI assistant to their website without hiring engineers or spending months integrating tools.
That's exactly why we built Fabric Applications, starting with AI Assistants. And we've just rolled out a major set of improvements that make them feel more natural, more engaging, and more helpful for your visitors. If your assistant is already live, these updates are available now.
The Problem With Website Chatbots Today
If you've ever looked into adding a chatbot to your website, you've probably heard some version of this:
- "It'll take a few weeks."
- "You'll need a developer."
- "We'll need to set up hosting, APIs, and monitoring."
- "It'll cost a few thousand dollars upfront."
For many companies, especially ecommerce brands, marketplaces, and startups, that's a non-starter.
The irony is that the use case is simple:
"Answer customer questions using our existing content."
But the implementation has historically been anything but simple.
What the Fabric AI Assistant Is
The Fabric AI Assistant lets you create, train, and deploy a chatbot in under 5 minutes.
- No engineering.
- No infrastructure setup.
- No ongoing maintenance.
You connect your content, customize how the assistant behaves, and embed it on your site with a single line of code. That's it.
How It Works (Without the Technical Stuff)
1. Add your content
You can connect:
- Your website
- Help docs or FAQs
- PDFs or internal documentation
Fabric reads and understands your content automatically, the same way a real customer would.
2. Customize the assistant
You decide:
- How it talks (friendly, professional, sales-focused)
- What it should answer
- What it should avoid
- When to suggest contacting a human
No prompts. No configuration files.
Want to try this yourself? Get started in minutes.
3. Deploy instantly
Copy one script tag. Paste it into your site. Your AI assistant is live.
What's New: A Smarter, More Interactive Experience
We've shipped a series of upgrades that transform the assistant from a simple Q&A tool into something that feels genuinely conversational and engaging. Here's what's changed.
Real-Time Streaming Responses
Responses now appear word by word instead of loading all at once. It feels faster, more conversational, and more human, even when the answer is detailed.
Built-In Feedback
Visitors can give a thumbs up or thumbs down after each reply, giving you a clear signal on what answers are working and where improvements might be needed.
Suggested Follow-Up Questions
After each response, the assistant can surface relevant follow-up questions that visitors can click. This guides people deeper into your product, keeps conversations moving, and helps users explore without having to think of the next question.
Voice Input
Visitors can tap a microphone button and speak instead of typing. This is especially useful for mobile users, accessibility, and situations where typing isn't convenient.
Voice Output
Responses can also be read aloud with a single click. Ideal for longer answers, on-the-go users, and anyone who benefits from audio accessibility.
Automatic Multilingual Support
If someone types in Spanish, French, or another language, the assistant replies in that same language automatically. No extra configuration required. This is a natural fit for ecommerce stores with global customers, international SaaS products, and public-facing organizations.
Proactive Help
You can now configure the assistant to appear automatically based on time spent on a page or scroll depth. Show it after a few seconds on pricing pages, during checkout, or when someone reaches the middle of a long product page or blog post. If a visitor opens the chat themselves first, it won't interrupt them.
Why This Is Different
This isn't the first chatbot in the world. What's different is how easy it is to use and how little overhead it requires.
Traditional chatbot tools rely on always-on infrastructure and custom setup. Fabric doesn't. The assistant runs only when it's needed, which means:
- Faster setup
- Lower cost
- No infrastructure to manage
For most teams, that changes the decision from "maybe later" to "why not now?"
What This Looks Like in Practice
If you're an ecommerce company
- Answer shipping, returns, and product questions instantly
- Help customers find the right product faster
- Reduce repetitive support tickets
- Turn questions into conversions
- Reach global customers with automatic multilingual replies
- Trigger proactive help on checkout pages to reduce abandonment
If you're a marketplace or SaaS product
- Explain features and pricing clearly
- Onboard users without extra support load
- Provide consistent answers at any time of day
- Use suggested follow-ups to guide users through complex workflows
- Collect feedback to continuously improve responses
If you're a research or data-driven startup
- Let users query documentation conversationally
- Turn dense PDFs into usable answers
- No ML team required
- Support voice input for hands-free research workflows
Setup Time vs. Reality
| With Fabric AI Assistants | |
|---|---|
| Setup time | Minutes, not months |
| Engineering required | None |
| Infrastructure | Included |
| Cost | Typically 70–90% lower |
| Deployment | One script tag |
| New features | Automatic, no action needed |
Why This Matters
Fabric Applications exist to make powerful systems usable by anyone.
The AI Assistant is no longer just a chatbot. With streaming responses, voice input and output, multilingual support, proactive engagement, and built-in feedback, it's become a genuinely interactive experience: one that's more accessible, more engaging, and better at guiding your customers.
- It's built on top of advanced infrastructure
- But you never have to think about that
- You just get the result
And it still takes just minutes to set up.
What Comes Next
The AI Assistant is just the first Fabric Application.
Going forward, we'll keep shipping application-level tools that combine multiple capabilities into simple, usable products, tools that used to require teams, servers, and long timelines.
Fabric isn't becoming a single product.
It's becoming a platform where practical, real-world tools can be created quickly, and used immediately by the people who need them.