Spotify’s new Natural Language API Interface and other Examples Explored
How the world’s top product teams are introducing natural language interfaces in their products. Canva, Slack, Hubspot, Linear, Shopify, Vercel and more.
🔒DoP Deep goes deeper into the concepts and ideas that are covered in the Weekly Briefing to help you learn lessons from the experiences of top tech companies. If you’d like to upgrade to receive these in-depth pieces of analysis you can upgrade below. New reports are added every month.
The ability to speak to products in natural language is transforming pretty much every part of the product experience.
We’ve come to expect natural language interfaces (NLIs) in user-facing features like AI Assistants, but now we’re also starting to see it roll out to other parts of the product stack, too.
For example, just this past week, Spotify revealed that it has built a new plugin that lets its advertisers create and and manage ad campaigns in the Spotify API using plain language instead of having to write manual API calls. As well as this, Uber’s CPO revealed new conversational features in their annual conference and one CEO of AI company LlamaIndex put it nicely this week when he commented on his own current experience with engineering teams:
“Engineers are not actually writing real code…They’re all typing in natural language”
In this Deep Dive, we’ll explore some real world examples of new natural language interfaces and how NLI is transforming the features we build, as well as the underlying technical infrastructure that powers them.
Coming up:
Spotify’s Natural Language ads builder and 30+ other examples of natural language interface UX in other products
How one company built an NLI feature with 85% adoption rates
From “Execution UI” to “Judgment UI” - a useful frame for deciding which screens in your product to invest in, simplify, or retire
Prompts you can use to identify which workflows in your own product might be ready for NLI
The full 30+ examples of NLI in action, from Canva, Adobe, Spotify, Amazon, Google, Figma, Shopify, Slack, Airbnb, Uber, Yelp and more
How this analysis is structured
There are over 30 different examples of natural language interface features included in this piece from companies like Amazon, Adobe, Uber, Canva and more. Each example includes a description of how it works along with links to find out more.
Each example is also categorized to demonstrate how NLIs are now transforming every part of the product stack, including:
User-Facing Product - NLI features built directly into consumer or business products that end users interact with. The user replaces clicking, filtering, and form-filling with plain language. Examples include Spotify’s playlist creation (”lo-fi beats for a late-night study session”), Airbnb’s accommodation search (”a cozy cabin near a lake for a family reunion”).
Internal Tool - NLI applied to internal business workflows - things that help employees rather than customers. The user here is typically a knowledge worker, analyst, or support agent. Examples include Slack’s natural language workplace search, Databricks’ pipeline generation, and Linear Agent handling ticket triage autonomously.
Developer Tool - NLI aimed at software developers specifically - either to accelerate coding or to abstract away technical complexity
Platform / Framework - NLI capabilities that other products are built on top of, rather than end-user features themselves. Microsoft’s NLWeb, Google’s FunctionGemma, Vercel’s json-render, and ChatGPT’s app integrations layer all sit here. These are the infrastructure layer that makes the other three categories possible.
What Spotify built: a natural language to API ad builder
Let’s start with a recent example from Spotify. Here’s a snapshot of what Spotify’s engineering teams built recently to allow advertisers to describe their campaigns in natural language:
