Moving from clicks to conversation with natural language commands
Overview
Why apply the filters yourself, when you can type in English related to what you’re looking for? The aim with this pattern isn’t to get you one final result, but to help you use the app’s UI using natural language. In the examples provided, it’s a search, but the command could apply to any goal a user is trying to achieve.
User intent
Getting result faster
Macro trend
AI in old UI
Why do we need natural language searching and filtering?
Humans think in sentences, not dropdowns, checkboxes, and toggles. We don't think "price: low to high." We think "I need affordable blue sneakers for my vacation."
Traditional interfaces force us to adapt to them. This creates frustration and requires manual work that shouldn't be necessary. Introducing: Natural language interfaces. They work like a personal technical assistant, making the system adapt to us.
We're moving beyond just improving buttons and fields. We're replacing them with more human interfaces where users can start working directly to get things done easier and faster.
This is a step towards assistive UX.
Let's explore how this works in practice.
Examples
Linear
Linear replaced complex filtering with natural language brilliance. Just type "open bugs with SLAs" and the filters get applied themselves instantly. A highly intuitive and intent-based UX.

New users get pre-made prompts while power users can directly type in the search. The responsive loading indicator keeps us informed while the system processes our request.

Stripe
docs uses the same conversational approach but for technical content. Ask "What is test mode?" and get precise answers pulled from their entire documentation. Technical docs finally feel natural instead of overwhelming.

Their "Gathering sources" loading state perfectly builds on user's anticipation. They mark AI responses as experimental to manage expectations and capture feedback with a simple, useful/not useful option (the standard).

Designing natural language input isn't about making things simpler. It's about understanding how people think. If we misinterpret what users mean, trust breaks down quickly.
The hard part? Making natural language feel flexible, not another syntax that you have to learn.

"I've gotten a ton of value out of aiverse over the last year!"
Dave Brown, Head of AI/ML at Amazon

Unlock this pattern
instantly with PRO
Access the entire Pattern Library
Access all upcoming Checklists
Access all upcoming Case studies
Get on-demand AI insights for your UX challenges
Curated by
